1234567 151 - 200 of 478
rss atomLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
  • Engström, Isak
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Ihre, Alan
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Predicting house prices with machine learning methods2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    In this study, the machine learning algorithms k-Nearest-Neighbours regression (k-NN) and Random Forest (RF) regression were used to predict house prices from a set of features in the Ames housing data set. The algorithms were selected from an assessment of previous research and the intent was to compare their relative performance at this task.

    Software implementations for the experiment were selected from the scikit-learn Python library and executed to calculate the error between the actual and predicted sales price using four different metrics. Hyperparameters for the algorithms used were optimally selected and the cleaned data set was split using five-fold cross-validation to reduce the risk of bias. An optimal subset of hyperparameters for the two algorithms was selected through the grid search algorithm for the best prediction.

    The Random Forest was found to consistently perform better than the kNN algorithm in terms of smaller errors and be better suited as a prediction model for the house price problem.

    With a mean absolute error of about 9 % from the mean price in the best case, the practical usefulness of the prediction is rather limited to making basic valuations.

  • Gessle, Gabriel
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Åkesson, Simon
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    A comparative analysis of CNN and LSTM for music genre classification2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    The music industry has seen a great influx of new channels to browse and distribute music. This does not come without drawbacks. As the data rapidly increases, manual curation becomes a much more difficult task. Audio files have a plethora of features that could be used to make parts of this process a lot easier. It is possible to extract these features, but the best way to handle these for different tasks is not always known. This thesis compares the two deep learning models, convolutional neural network (CNN) and long short-term memory (LSTM), for music genre classification when trained using mel-frequency cepstral coefficients (MFCCs) in hopes of making audio data as useful as possible for future usage. These models were tested on two different datasets, GTZAN and FMA, and the results show that the CNN had a 56.0% and 50.5% prediction accuracy, respectively. This outperformed the LSTM model that instead achieved a 42.0% and 33.5% prediction accuracy.

  • Engelhardt, Erik
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Jäger, Simon
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    An evaluation of image preprocessing for classification of Malaria parasitization using convolutional neural networks2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    In this study, the impact of multiple image preprocessing methods on Convolutional Neural Networks (CNN) was studied. Metrics such as accuracy, precision, recall and F1-score (Hossin et al. 2011) were evaluated. Specifically, this study is geared towards malaria classification using the data set made available by the U.S. National Library of Medicine (Malaria Datasets n.d.). This data set contains images of thin blood smears, where uninfected and parasitized blood cells have been segmented.

    In the study, 3 CNN models were proposed for the parasitization classification task. Each model was trained on the original data set and 4 preprocessed data sets. The preprocessing methods used to create the 4 data sets were grayscale, normalization, histogram equalization and contrast limited adaptive histogram equalization (CLAHE).

    The impact of CLAHE preprocessing yielded a 1.46% (model 1) and 0.61% (model 2) improvement over the original data set, in terms of F1-score. One model (model 3) provided inconclusive results. The results show that CNN’s can be used for parasitization classification, but the impact of preprocessing is limited.

  • Bodelsson, Andrea
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Kollektiv odling i kris: - en antropocen trädgård2019Independent thesis Basic level (professional degree), 20 poäng / 30 hpOppgave
    Abstract [sv]

      “Kollektiv odling i kris - en antropocen trädgård"  visar en vision av hur ytor i en svensk stad har omformats och omplanerats, kanske som en konsekvens av samhälleliga beslut, kanske genom ideellt arbete - kanske bäggedera.   Bakgrunden till allt det här är såklart klimatförändringarna, som med stor sannolikhet om några år även kommer påverka matförsörjningen. Det blir då nödvändigt att odla mer lokalt, i Sverige och i våra städer där merparten bor och lever.   Idén är att odlingsmarken och växthus genererar grönsaker och baljväxter, som vi idag importerar, och som kommer att utgöra en större andel av vår kost i framtiden. Transporterna blir minimalt korta, och maten konsumeras i första hand av stadsborna runt omkring, samtidigt som avfallet från odlingen genererar värme och jord till växthuset och odlingen.   Runt varje trädgård går en mur av jord som skyddar odlingen från ljud och avgaser från motorvägen, och även från vinden. Muren blir en tydlig figur i stadslandskapet som ramar in och skyddar odlingen. Alla trädgårdar har en mittaxel där vattnet leds, och bredvid ligger odlingarna, som ramas in av jordväggar som värmer upp jorden. Strukturen förstärks av längor av växthus längs med jordväggarna.   Den antropocena trädgården är varken en utopi eller dystopi. Projektet handlar bara om hur framtiden kanske helt nödvändigt kan komma att se ut, och om hur staden kan planera för att möta framtida behov.

  • Holmgren Faghihi, Josef
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Gorgis, Paul
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Time efficiency and mistake rates for online learning algorithms: A comparison between Online Gradient Descent and Second Order Perceptron algorithm and their performance on two different data sets2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    This dissertation investigates the differences between two different online learning algorithms: Online Gradient Descent (OGD) and Second-Order Perceptron (SOP) algorithm, and how well they perform on different data sets in terms of mistake rate, time cost and number of updates. By studying different online learning algorithms and how they perform in different environments will help understand and develop new strategies to handle further online learning tasks. The study includes two different data sets, Pima Indians Diabetes and Mushroom, together with the LIBOL library for testing. The results in this dissertation show that Online Gradient Descent performs overall better concerning the tested data sets. In the first data set, Online Gradient Descent recorded a notably lower mistake rate. For the second data set, although it recorded a slightly higher mistake rate, the algorithm was remarkably more time efficient compared to Second-Order Perceptron. Future work would include a wider range of testing with more, and different, data sets as well as other relative algorithms. This will lead to better result and higher credibility.

  • Bergström, Mattias
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Fahlander, Per
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Evaluating a Novel, Scalable Natural Language Processing Heuristic for Determining Semantic Relatedness2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

     Distributional semantics is a recent research field aiming to quantify how close one text is to another in terms of contextual meaning. In this study we propose and evaluate a novel distributional semantics model on how much agreement its predictions can yield with a set of 12,227 human opinions. We call this method Refined Semantic Relatedness (RSR), which applies an incrementally improvable word association index and some distributional principles for producing theoretically educated predictions. Using 1951 preprocessed Wikipedia articles as a basis for the predictions, the model predicted the human opinions with a Pearson correlation of 0.3. In previous literature it has been claimed that Explicit Semantic Analysis (ESA-Wiki) achieve a corresponding Pearson correlation of 0.72 by utilizing 241,393 preprocessed Wikipedia articles. That is roughly 5.76 times more variance accounted for, although, also a result of considerably more extensive preprocessing in terms of articles. While the predictive value of RSR turned out relatively low as a result of the study’s limitations, this could be addressed in further research. We believe that this paper in any way can contribute with some novel ideas to the field.

  • Bernerstedt, Ingrid
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Varvet: A concert venue on Södermalm2019Independent thesis Basic level (professional degree), 20 poäng / 30 hpOppgave
    Abstract [en]

    This project proposes a new building for a concert venue on Södermalm in Stockholm. An area especially affected by the current trend that venues for experiencing contemporary music live are being shut down and driven away from the inner city, due to everything from complaining neighbours to making room for more profitable businesses. The idea for the project sprung out of the sudden closing of the venue Debaser Medis, whose facilities was removed when the building it was situated in was to undergo major renovations. The purpose with Varvet was to create a new venue in a freestanding building, independent from other businesses, and giving contemporary music a more articulated place in the city. The project also includes the re-purposing of an adjacent existing building to facilitate a cafe and bar as well as functions supporting the daily business of the venue such as offices, practice rooms and a small workshop. 

  • Taherifard, Ershad
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Load and Demand Forecasting in Iraqi Kurdistan using Time series modelling2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    This thesis examines the concept of time series forecasting. More specifically, it predicts the load and power demand in Sulaymaniyah, Iraqi Kurdistan, who are today experiencing frequent power shortages. This study applies a commonly used time series model, the autoregressive integrated moving average model, which is compared to the naïve method. Several key model properties are inspected to evaluate model accuracy. The model is then used to forecast the load and the demand on a daily, weekly and monthly basis. The forecasts are evaluated by examining the residual metrics. Furthermore, the quantitative results and the answers collected from interviews are used as a basis to investigate the conditions of capacity planning in order to determine a suitable strategy to minimize the unserved power demand. The findings indicate an unsustainable over consumption of power in the region due to low tariffs and subsidized energy. A suggested solution is to manage power demand by implementing better strategies such as increasing tariffs and to use demand forecast to supply power accordingly. The monthly supply forecast in this study outperforms the baseline method but not the demand forecast. On weekly basis, both the load and the demand models underperform. The performance of the daily forecasts performs equally or worse than the baseline. Overall, the supply predictions are more precise than the demand predictions. However, there is room for improvement regarding the forecasts. For instance, better model selection and data preparation can result in more accurate forecasts.

  • Axelsson, Henrik
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Wass, Daniel
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Machine Learning for Activity Recognition of Dumpers2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    The construction industry has lagged behind other industries in productivity growth rate. Earth-moving sites, and other practices where dumpers are used, are no exceptions. Such projects lack convenient and accurate solutions for utilization mapping and tracking of mass flows, which both currently and mainly rely on manual activity tracking. This study intends to provide insights of how autonomous systems for activity tracking of dumpers can contribute to the productivity at earthmoving sites. Autonomous systems available on the market are not implementable to dumper fleets of various manufacturers and model year, whereas this study examines the possibilities of using activity recognition by machine learning for a system based on smartphones mounted in the driver’s cabin. Three machine learning algorithms (naive Bayes, random forest and feed-forward backpropagation neural network) are trained and tested on data collected by smartphone sensors. Conclusions are that machine learning models, and particularly the neural network and random forest algorithms, trained on data from a standard smartphone, are able to estimate a dumper’s activities at a high degree of certainty. Finally, a market analysis is presented, identifying the innovation opportunity for a potential end-product as high.

  • Olson, Julius
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Södergren, Emma
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Long Term Memory in Conversational Robots2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    This study discusses an implementation of a long term memory in the robot Furhat. The idea was to find a way to prevent identical and very similar questions from being asked several times and to store the information of which questions have already been asked in a document database. The project encompasses tf-idf, as well as a small-scale test with Word2Vec, to find a vector representation of all questions from Furhat’s database and then clustering these questions with the k-means method. The tests resulted in high scores on all the evaluation metrics used, which is promising for implementation into the actual Furhat robot, as well as further research on similar implementations of long term memory functions in chatbots.

  • Bergström, Emelie
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Durowicz, Aleks
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Establishing requirements for a gamified digital math tool in a South African township primary school: A case study based on gamification and interaction design2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

     This project is based on the desire to increase the understanding and knowledge of digital resource usage within the math education in South African townships. Digital education is considered to facilitate solutions for current shortcomings, yet knowledge and experience of the local circumstances are vital for successful implementation. Interviews, observations and analyses of existing games were used to define requirements regarding design and usability, which were implemented through prototypes and evaluated formatively. This study indicates that the requirements of an educational and stimulating digital tool are the most important and suggest consideration of feedback, gamification features, such as storytelling, as well as the balance of formal and informal settings. Furthermore, user characteristics heavily implies low complexity, few distractions and a logical structure. Requirements of appropriate level, in both language and content, also proved to be crucial. However, this demands further assessment before implementation. The developed tool was perceived as suitable and was accepted as a long-term solution to make better use of the current resources. Although the limited time and the difficulty to comprehend the users’ opinions bounded the conclusions, the established requirements are suggested to provide a solid foundation for further research.

  • Gonzalez Munoz, Mario
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Hedström, Philip
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Predicting Customer Behavior in E-commerce using Machine Learning2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [sv]

    E-handel har varit en snabbt växande sektor de senaste åren och förväntas fortsätta växa i samma takt under de närmsta. Detta har öppnat upp nya möjligheter för företag som försöker sälja sina produkter och tjänster, men det tvingar dem även att utnyttja dessa möjligheter för att vara konkurrenskraftiga. En intressant möjlighet som vi har valt att fokusera detta arbete på är förmågan att använda kunddata, som inte varit tillgänglig i fysiska butiker, till att identifiera mönster i kundbeteenden. Förhoppningsvis ger detta en ökad förståelse för kunderna och gör det möjligt att förutspå framtida beteenden. Vi fokuserade specifikt på att skilja mellan potentiella köpare och faktiska köpare, med avsikt att identifiera nyckelfaktorer som avgör ifall en kund genomför ett köp eller ej. Detta gjorde vi genom att använda Binary Logistic Regression, en algoritm som använder övervakad maskininlärning för att klassificera en observation mellan två klasser. Vi lyckades ta fram en modell som förutsåg om en kund skulle genomföra ett köp eller ej med en noggrannhet på 88%.

  • Kästel, Arne Morten
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Vestergaard, Christian
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Comparing performance of K-Means and DBSCAN on customer support queries2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    In customer support, there are often a lot of repeat questions, and questions that does not need novel answers. In a quest to increase the productivity in the question answering task within any business, there is an apparent room for automatic answering to take on some of the workload of customer support functions. We look at clustering corpora of older queries and texts as a method for identifying groups of semantically similar questions and texts that would allow a system to identify new queries that fit a specific cluster to receive a connected, automatic response. The approach compares the performance of K-means and density-based clustering algorithms on three different corpora using document embeddings encoded with BERT. We also discuss the digital transformation process, why companies are unsuccessful in their implementation as well as the possible room for a new more iterative model.

  • Mekonnen, Michael
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Tahir, Gara
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Vilka kriterier är viktiga för användarupplevelsen vid interaktion med en språkcafé-robot?2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    As the number of immigrants in Sweden rises, the demand for alternative methods for language learning increases simultaneously. The use of social robots for teaching a second language is a promising field. The following research question has been designed to identify how social robots can be improved to better suit second language learners. The research question is: Which criteria are important for the user experience when interacting with a language cafe robot? The main method used to answer the question is Design Thinking with the help of semi-structured interviews. The results were 12 criteria which can be implemented for social robots in the future. The research has also studied how the criteria can be implemented in robots and to what degree the robot Furhat developed by Furhat robotics has implemented the criteria today.

  • Duque, Fredd
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Six DOF tracking system based on smartphones internal sensors for standalone mobile VR2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Nowadays mid-range smartphones have enough computational power to run simultaneous location and mapping (SLAM) algorithms that, together with their onboard inertial sensors makes them capable of position and rotation tracking. Based on this, Google and Apple have released their own respective software development kits (SDKs) that allow smartphones to run augmented reality applications using six degrees of freedom tracking. However, this same approach could be implemented to virtual reality head-mounted-display (HMD) based on smartphones, but current virtual reality SDKs only offer rotational tracking. In this study the positional tracking technology used for augmented reality mobile applications has been implemented in a virtual reality head-mounted-display only powered by a smartphone by combining virtual and augmented reality SDKs. Compatibility issues between SDKs have been faced to develop a working prototype. An objective and controlled measurement study has been conducted that included 34.200 measurements, to test the accuracy, precision and jitter tracking of the protype against the Oculus Rift, a dedicated virtual reality system. Results show that the developed prototype offers a decent tracking precision and accuracy in optimal conditions. It was concluded to be highly dependent on the camera view. Although, jitter presented the opposite behavior, being dependent to the device used but independent on the camera view. In its optimal conditions, user studies demonstrated that the prototype was capable of offering the same tracking performance feeling as the Oculus Rift although jitter was quite noticeable, and a common user complain. Further studies are proposed that can improve the tracking performance of the prototype by filtering jitter and using two or more cameras with a different angular to correlate feature points and obtain a wider view of the environment were the prototype is used.

  • Lindström, Erik
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Animation of humanoid characters using reinforcement learning2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Procedural animations are still in its infancy, and one of the techniques to create such is using Reinforcement Learning. In this project, swimming animations are created using UnityML version 0.6 with their Reinforcement Learning training agents, using the policy PPO, created by OpenAI. A humanoid character is placed in a simulated water environment and propels itself forward by rotating its joints. The force created depends on the joints mass and the scale of the rotation. The animation is then compared to a swimming animation created using movement capture data. It is concluded that the movement capture data animation is significantly more realistic than the one created in this project. The procedurally created animations displays many of the typical issues with reinforcement learning such as jittering and non-smooth motions. While the model is relatively simple, it is not possible to avoid these issues completely with more computational power in the form of a more complex model with more Degrees of Freedom. It is however possible to finetune the animations with the improvements listed at the end of the discussion.

  • Lönn, Caroline
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Augmented Reality Smartphone Applications as a Tool to Raise Awareness of Circular Economy2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Environmental problems are increasingly jeopardizing the earth's life-support systems. By shifting from a take-make-dispose industrial model to a circular one, resources will last longer, and the environmental impact will be significantly lower.

    People’s choices, behaviours and lifestyles will play a vital role in achieving sustainable development. A way to influence this is to use an educational tool that is accessible, interesting and fun. By superimposing computer generated objects to the real world, Augmented Reality (AR) adds a layer of interactivity and engagement. As there are 3 billion active smartphone users worldwide, an AR smartphone application is also very accessible.

    This thesis investigates how marker-based AR can be used as a means to educate a user in circular economy. This was done by using research through design and applied research to create a design for a prototype which worked as a proof of concept. The prototype was evaluated through a heuristic evaluation. Data was gathered through a form and semi-structured interviews. The data was analysed using the framework method and box plots.

    The prototype used marker-based AR and was built using the game engine Unity with the Vuforia SDK. The results indicated that an AR application could be a good tool to use in combination with other sources of information such as seminars.

  • Olofsson, Oskar
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Detecting Unsynchronized Audio and Subtitles using Machine Learning2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Unsynchronized audio and subtitle files are common within streaming media. As subtitles often are an essential part of the viewing experience, this can have large consequences, possibly making the content inaccessible. Detecting the unsynchronization manually is a time consuming task, as entire media files have to be viewed and evaluated by a person.

    In this thesis an investigation on how to detect unsynchronized audio and subtitles automatically using machine learning is performed. The process is divided into two parts. The first part consists of training the models Support Vector Machine, Random Forest and Multilayer Perceptron to classify whether subtitles should be present given features extracted from audio. As a part of this process the algorithms are compared and evaluated based on their accuracy and time-efficiency. The second part is composed of using the best model to detect unsynchronization. It is done through a similarity measurement between the predicted subtitle distribution and the distribution of the actual subtitles. If a better similarity can be found through shifting the subtitles, the files are classified as unsynchronized.

    The project shows that Random Forest has the highest accuracy and is thus best suited for the purpose. Of ten file pairs tested for unsynchronization the method successfully categorized nine of them. The conclusion is that the approach is working, yet future work includes increasing the accuracy through testing other algorithms and audio feature extraction techniques.

  • Skeppstedt, David
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Identification and Exploitation of Vulnerabilities in a Large-Scale ITSystem2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    This thesis presents the results of a vulnerability assessment and exploit development targeting a large-scale IT-system. Penetration testing and threat modelling was used to identify vulnerabilities in the system. This resulted in identification of five vulnerabilities and the development of a reliable denial of service exploit using an authentication bypass and a stack-based buffer overflow. The consequences of the vulnerabilities and the exploit is discussed and set into a broader perspective. The conclusion is that the results from this thesis can help improve the security of the IT-system. However, the identification of additional vulnerabilities could lead to a more potent exploit.

  • Jefford-Baker, Jonathan
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    ALCOL: Probabilistic Threat Modelling of the Amazon Elastic Container Service Domain2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Cloud computing is becoming an increasingly popular computation model for IT-infrastructures which has changed the notion of computing resources. Another concept that has become popular is containers which provides the capability to run applications isolated from each other while sharing the host’s operating system kernel. These two concepts have been combined to run containerised environments in the cloud, a cloud service type which has become popular among customers. The increased deployment of IT-infrastructures built on cloud environments running containers results in an increased exposure to cyber attacks within this domain which requires that proper security measures are taken.

    Assessing the security of a system can, however, be difficult. Attack simulations can be used to provide an overview of how an adversary can attack the system to simplify this task. This thesis proposes a probabilistic threat modelling language which can be used to simulate attacks against infrastructures based on Amazon Elastic Container Service (ECS), a cloud service provided by Amazon Web Services which allow customers to run containerised applications in the cloud. The language, called ALCOL (Amazon eLastic COntainer Language), is based on the Meta Attack Language and the domain-specific language AWSLang.

    The language was developed using multiple literature studies to discover the different components in Amazon ECS that should be modelled in the language, as well as the different attacks possible to perform against Amazon ECS infrastructures. The language was evaluated using test cases representing different attack scenarios and also through an interview with a domain expert.

    The developed language is able to accurately simulate cyber attacks against Amazon ECS infrastructures, although with some limitations, which lead to propositions for future research.

  • Ringh, Matilda
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Gnesta Community Center2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    Gnesta community center is located at the main square, right next to the railway station. As the city grows, a new center block is planned where the community center would be located on the eastern end of the block. My ambition was to create a building that reflects the small-town character found in Gnestaby working with materials from the site, intimate rooms and meetings with the citizen. By making use of the material from the old station house across the street, the buildings create a portico to the city. The meeting between the railway, the station building, the main square and the community center becomes a clear centerpiece in the city and a representative place for Gnesta.

  • Cata Villa, Marcel
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    3D Bounding Box Detection from Monocular Images2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Object detection is particularly important in robotic applications that require interaction with the environment. Although 2D object detection methods obtain accurate results, these are not enough to provide a complete description of the 3D scenario. Therefore, many models have recently showed promising progress in this challenging field.

    In this work, the goal is to predict 3D bounding boxes from single images without using temporal data or any explicit depth estimation. We propose an approach for 3D monocular object detection based on Deep3DBox. We replace the geometric constraints taken into account to predict the 3D location of objects by a deep learning module. Moreover, we undertake a study on the different parameters for the modules that are used to predict dimensions and orientation of objects.

    We conduct experiments in order to search for the best hyperparameters of our model for KITTI cars and we report and compare our results on KITTI and the challenging NuScenes benchmarks for cars and pedestrians with other state of the art methods. Therefore, we conclude that our approach performs on par with similar methods and improves Deep3DBox results.

  • Arendal Rundbom, Vera
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Town hall in Nynäshamn2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    Stadshusplatsen in Nynäshamn is a square with good prerequisites to become an oasis for the citys inhabitants. Though today the current town hall, a low-rise building, closes the square towards the south. The old town hall has lost its function as a uniting building and together with Folkets Hus, they form a wall that blocks the sun and the view towards the sea.

    The assignment was to draw a proposal for a new town hall where the current town hall is situated and to explore how the new structure can relate to an adjacent high-rise building and create new dimensions and prerequisites to the square.

    First of all, I chose to create a passage between Folkets Hus and the new town hall. This to let the sun shine in on the square and create a connection between the square and what is today the back of the town hall but has potential of becoming a sunny meeting point for the citizens of Nynäshamn.

    To find ways to create a connection in terms of design between the high-rise building and my proposal I looked at the base floor of the high-rise building, which consists of a pillar arcade. Based on this arcade two grids, one primary and one secondary, arise and creates the conditions of the building both spatially and structurally. In the intersections of the two grids I’ve placed pillars, one larger and one smaller type, to define spatialities both outside and inside the building.

    The two different pillar types also help to define in which zone of the building you are located. The smaller pillars represent a more intimate zone and the larger ones the public zones. In the more intimate zone, two cores of service have been placed. Furthermore, no fixed walls are used but the rooms are defined by columns and three different wall types: a light wall, a flexible sliding wall and a movable fabric wall.

  • Freimanis, Davis
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Vulnerability Assessment of Authentication Methods in a Large-Scale Computer System2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Vulnerabilities exist in almost all software programs. Some software is more vulnerable than others. A method that can be used to mitigate the vulnerabilities is penetration testing. In this thesis, a penetration test was conducted on a large scale computer system provided by a company. The goal of the thesis was to see if vulnerabilities could be found, with a focus on the field of authentication. After conduction a thorough penetration test there were vulnerabilities found that threaten the confidentiality and integrity of the system. Authentication vulnerabilities were found by leaking password hashes and by performing pass-the-hash and pass-the-ticket exploits.

  • Matsoukas, Christos
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Model Distillation for Deep-Learning-Based Gaze Estimation2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    With the recent advances in deep learning, the gaze estimation models reached new levels, in terms of predictive accuracy, that could not be achieved with older techniques. Nevertheless, deep learning consists of computationally and memory expensive algorithms that do not allow their integration for embedded systems. This work aims to tackle this problem by boosting the predictive power of small networks using a model compression method called "distillation". Under the concept of distillation, we introduce an additional term to the compressed model’s total loss which is a bounding term between the compressed model (the student) and a powerful one (the teacher). We show that the distillation method introduces to the compressed model something more than noise. That is, the teacher’s inductive bias which helps the student to reach a better optimum due to the adaptive error deduction. Furthermore, we show that the MobileNet family exhibits unstable training phases and we report that the distilled MobileNet25 slightly outperformed the MobileNet50. Moreover, we try newly proposed training schemes to increase the predictive power of small and thin networks and we infer that extremely thin architectures are hard to train. Finally, we propose a new training scheme based on the hintlearning method and we show that this technique helps the thin MobileNets to gain stability and predictive power.

  • Xiang, Ziyi
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    A comparison of genetic algorithm and reinforcement learning for autonomous driving2019Independent thesis Basic level (degree of Bachelor), 10 poäng / 15 hpOppgave
    Abstract [en]

    This paper compares two different methods, reinforcement learning and genetic algorithm for designing autonomous cars’ control system in a dynamic environment.

    The research problem could be formulated as such: How is the learning efficiency compared between reinforcement learning and genetic algorithm on autonomous navigation through a dynamic environment?

    In conclusion, the genetic algorithm outperforms the reinforcement learning on mean learning time, despite the fact that the prior shows a large variance, i.e. genetic algorithm provide a better learning efficiency.

  • Welin-Berger, Robert
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Return barriers and their application to stack tracing on modern VMs2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Virtual machines are today at the core of a very large portion of the code that runs our everyday lives. Their relative performance and characteristics are of high significance for society indirectly and few other are as impactful as the JVM. In this thesis, we examine the possibility to speed up stack tracing on HotSpot and OpenJDK by implementing a return barrier. The barrier is implemented by overwriting the return program counter on the top frame with a function that re-installs itself on the frame below every time it’s executed. Because of this, the barrier acts as a “low water mark” and we then leverage that fact to enable Java Flight Recorder to cache the section of the stack that has already been traversed before. While the first implementation of such a cache did not prove successful in terms of performance improvement we look at costs and benefits of different steps in the implementation and highlight where future use the barrier might be attempted.

  • Wiklund, Hannes
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Gnesta Community Center2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [sv]

    Ett fiktivt förslag för ett medborgarhus i Gnesta. Tomten ges av en planerad utveckling av Gnesta Centrum.Min byggnad har som mål att bli ett levande medborgarcenter som kommunicerar sin funktion genom sin helt transparenta fasad.

    Mer information på sid. 5.

  • Brink, Pontus
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Dissection of a Generative Network for Music Composition2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Controlling what a neural network generates has had great success when applied in the image domain. This thesis explores the performance of similar methods but instead applied in music generation. WaveNET, a state of the art neural network in audio synthesis and generation is trained using Generative Adversarial Networks to produce piano music. Two different methods for controlling the generation are presented, named HARD and SOFT. The HARD method fails to produce music of the same quality as without control. The SOFT method generates music of the same perceptual quality as without control but fails to control the output of the network. Finally, a discussion why this might be, and ideas regarding other methods for controlling the generation of music, and sequences in general are presented.

  • Östblom, Linnea
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Proposal for a new townhouse in Nynäshamn2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [sv]

    Förslag till nytt kommunhus i Nynäshamn

  • Disputas: 2019-10-28 14:00 F3, Stockholm
    Pechsiri, Joseph
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Hållbar utveckling, miljövetenskap och teknik.
    Nutrient Recovery as an Added Benefit to Harvests of Photosynthetic Marine Biomass: A Holistic Systems Perspective on Harvesting Marine Microalgae, Cyanobacteria, and Macroalgae2019Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    As a result of increasing environmental burdens from anthropogenic activities andresource scarcity, interest for the development of solutions utilizing photosyntheticmarine biomass has also been increasing in both academia and industries. Medium tolarge scale production and harvest of photosynthetic marine biomass have beenpracticed to achieve numerous services, including improving tourism industries,production of biofuels, and production of food/feed. However, few studies haveevaluated the potential for nutrient recovery as an added benefit to the aforementionedservices and the potential environmental burdens of such solutions from a holisticsystems perspective. This thesis, therefore, sought to determine the nutrient recoverypotential of harvesting photosynthetic marine biomass at industrial scales whileassessing the environmental burdens from a holistic systems perspective. Techniquesinvolving life cycle inventory and analysis, input-output analysis, growth modellingand experimentation, energy analysis, and assessment of greenhouse gas emissionsfrom a life cycle perspective were used to assess the potential environmental burdensof large scale harvest of photosynthetic marine biomass.This study employed five real world case studies of five different photosynthetic marinebiomass species at various geographical locations across the globe. Each case wasassessed to determine the potential to recover nutrients while evaluating the potentialenvironmental burdens from an energy and greenhouse gas perspective. Each casecontains unique specific details and therefore methods applied were case specific.Results showed that nutrient recovery potential existed in most cases with the exceptionof one case. Cases evaluated for their potential environmental burdens showed thatlarge scale harvest of photosynthetic marine biomass is resource intensive regardless ofspecies but showed mixed results from an energy perspective. The key findings of thisthesis were that a) the potential for nutrient recovery was estimated in both large scalecultivation and large scale wild harvest of photosynthetic marine biomass, b) from anenergy and biomass harvesting perspective, the viability of industrial harvests ofphotosynthetic marine biomass were found for both large scale cultivations and wildharvesting of biomass blooms, and c) scale of operations is an important factor towardsevaluating the environmental performance of photosynthetic marine biomassproduction systems.

  • Coronado Pons, Mar
    KTH, Skolan för industriell teknik och management (ITM).
    Modelling the performance of heat pump systems for single-family house applications2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Ground source heat pumps (GSHP) extracting the heat through borehole heat exchangers (BHEs) are an extremely efficient way to provide heating. Their performance is affected by the temperature of the thermal source: the ground; the higher it is the temperature of the ground, the higher their performance. As the demand of this heating technology increases, the amount of GSHP in densely populated areas is at risk of escalating notably. Consequently the study of thermal influence between neighbouring GSHPs is of paramount importance to properly design these systems in such areas.

    A comparison is made between the performance of an isolated house, and the same house as part of an area with high density of houses using identical GSHPs. The aim of the project is to study the long term consequences of exploiting the ground thermal source in an extensive manner, to analyse how the GSHP operation is affected in this specific case study, and present a methodology general enough to be implemented for different conditions.

    It is presented a methodology based on a parameter calibration model for the HP to analyse the performance along the years of a ground source heat pump system located in an area where there is a high density of identical installations. The model was tested to verify its accuracy when simulating the performance of the HP and was implemented for two case studies that emulate the conditions found in Sweden for residential heating. For the first case study, where a 6kW HP unit is simulated, the COP of the system decreased around 15% for the 25 studied years. In good agreement with this decline of the COP, an electricity consumption increase above 10% is faced. For the second case study, a heat pump unit double the size of the one employed for the first case is modelled. In this case, the drop for the COP is 16% and the electricity consumption growth is above 20%.

  • Markström, Ingemar
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Comparing normal estimation methods for the rendering of unorganized point clouds2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Surface normals are fundamental in computer graphics applications such as computer vision, object recognition, and lighting calculations. When working with unorganized point clouds of surfaces, there exists a need for fast and accurate normal estimation methods.

    This thesis presents the investigation and implementation of two different methods of normal estimation on fixed-size local neighborhoods in unorganized pointclouds.

    Two main categories of tests were conducted. The first type was visual inspection and the second consisted of numeric analysis of the normal estimation process and results. Point cloud data used in the study included numerically exact representations of spheres, cubes, cones, as well as both uniformly sampled or laser-scanned real-world point clouds with millions of points.

    Complete triangle averaging was found to be the method of choice on small neighborhoods, justified by faster running-time while still estimating high-quality normals. When larger neighborhood sizes were needed, a size breakpoint was found above which principal component analysis should be used instead, which estimates normals of similar quality as the complete triangle averaging but with the added benefit of near-constant running-time independent of neighborhood size.

  • Ala-Mutka, Jonatan
    KTH, Skolan för industriell teknik och management (ITM).
    Powering Africa by Empowering its People: An Action Research study at a Zambian microgrid company building local capacity to reach large scale viability2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Despite recent advances in the global electrification rates, increasing from 76% in 1990 to 85% in 2012, the United Nations goal of universal access to electricity by 2030 is still far from achieved, with an estimated 1.1 billion people still without access to electricity. Over half of these live in Sub-Saharan Africa, with a majority in rural areas and extreme poverty. Major challenges are inert with the current electrification path of centralized grid extension, leaving these people without power in decades to come. Microgrids, a decentralized power system consisting of solar power generation, energy storage and distribution technology, has been hailed as the only option to provide life improving and productivity inducing power for rural communities in Africa.

    However, despite recent hype and development in the sector, the diffusion of microgrids is still incremental due to a lack of viable large-scale operation, required for profitability. This is explained by targeting customers in remote rural areas with low ability to pay, and the task of delivering expensive technology and complex operations needed to manage and operate the grids. No industry blueprint or research on how to operate microgrids at scale or profitably exists. This thesis explores one blueprint, with the promise to increase profitability and allow for a more sustainable scaling. Local Capacity building is a decentralized approach by developing capacity directly in the local communities, through recruiting, skills development and training of people to be employed to operate and manage their local microgrids.

    The results consist of a framework outlining what local capacity building is, through research propositions that define the key components capturing the complete system of local capacity building is for scaling a microgrid business, along with the challenges and opportunities associated with scaling a business using local capacity building. It has been developed iteratively by application of an action research approach conducted on a small-scale Zambian Microgrid company facing radical growth. The researcher was immersed in the context, at the heart of this change, and in a participatory and interventionist fashion turning every stone to explore what local capacity building is, resulting in a robust study anchored in the field. Because of the contextually embedded nature of the data, this also means that the results are local. It is up to the reader to assess the applicability of the results in another context.

    The extensive results span multiple areas of the business, capturing the complexity of local capacity building, and contribute to knowledge on a holistic level on what local capacity building is. This blueprint was deemed viable to further develop in the small-scale Zambian microgrid company, specifically because of its potential to lower operating expenses and offer a more sustainable way to scale, and in extension diffuse microgrids in Africa.

  • Stiff, Harald
    KTH, Skolan för elektroteknik och datavetenskap (EECS).
    Explainable AI as a Defence Mechanism for Adversarial Examples2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Deep learning is the gold standard for image classification tasks. With its introduction came many impressive improvements in computer vision outperforming all of the earlier machine learning models. However, in contrast to the success it has been shown that deep neural networks are easily fooled by adversarial examples, data that have been modified slightly to cause the neural networks to make incorrect classifications. This significant disadvantage has caused an increased doubt in neural networks and it has been questioned whether or not they are safe to use in practice. In this thesis we propose a new defence mechanism against adversarial examples that utilizes the explainable AI metrics of neural network predictions to filter out adversarial examples prior to model interference. We evaluate the filters against various attacks and models targeted at the MNIST, Fashion-MNIST, and Cifar10 datasets. The results show that the filters can detect adversarial examples constructed with regular attacks but that they are not robust against adaptive attacks that specifically utilizes the architecture of the defence mechanism.

  • Jóhannsdóttir, Steinunn Kristín
    KTH, Skolan för kemi, bioteknologi och hälsa (CBH).
    Evaluation of Head and Neck Injuries during Misuses of Child Restraint Systems: Simulations of Car Accidents Performed with the PIPER Child Model2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Car collisions are, unfortunately, not uncommon and cause 1.35 million deaths each year worldwide. Children are often occupants in cars and to ensure their safety, child restraint systems (CRSs) have been developed. However, CRSs need to be used correctly to be efficient. Several studies, such as field investigations and Q-dummy tests, have shown that a misuse of a CRS can increase the risk of injuries.

    Typical misuses for a forward-facing CRS and a booster seat, with two real accident parameters, were constructed and simulated using the PIPER child human body model. The kinematics of each case were compared with injury parameters of the head, neck and abdomen. Comparing the parameters to existing injury criteria showed that most of the cases end in AIS3+ head injury, even cases with no misuse.

    When comparing the results of misuses to the cases where the CRS was correctly used, the dominant result was that misuse resulted in being less effective to protect the child. Moreover, results of chosen misuses compared to Q-dummy tests correlated with their results. Results from this thesis illustrate how important it is for parents to restrain children and route the belt correctly.

  • Bernhard, Isabelle
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Strängnäs City Hall2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    In my proposal of the new City Hall in Strängnäs there is several things I want to take advantage of in the project and for the estetics of the building;- The building represent some of old crafts of the city.- The silhuette of the building is accented with the crossing of the lines in the central staircase.- A clear connection of the visible floors of the City Hall to the next building (the neighbour bank).- Finding the driving components in the configuration of structural principles, the choosing of materials and the relations to the neighbourhood.- Binding the concept/design by different tests of rooms and program.

  • Blomberg, Gabriel
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Nynäshamn City Hall2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    A city hall is more than a building; it is a symbol that represents the ambitions of the city, a working democracy and civic interaction with politicians.

    The city hall must not only be welcoming for visitors and comfortable for workers, but also feel solid and durable to succesfully function as a symbol for the community for many years.

    Strong but welcoming, open but solid.

  • Bos, Anastasia
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Strängnäs City Hall2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [sv]

    Detta projekt är ett förslag på en ny stadshall i den mindre kommunen Strängnäs, belägen vid Mälaren. Området präglas av Strängnäs historia samt det varierande landskapet. Den givna platsen för nya stadshallen ligger vid Västervikstorget med båthamnen intill. På ena sidan ser man små äldre trähus växa upp för kullen med den gamla mjölkvarnen på toppen. På motsatta sidan syns murade och putsade kvartershus med domkyrkan i bakgrunden. Mötet mellan de två olika typologierna är spännande och även om denna typ av kontrast är vanlig i Strängnäs så framträder den starkast på platsen för den nya stadshallen.

    Den nya stadshallen är inspirerad av områdets karaktär med den varierande topografin och mötet mellan byggnadstypologierna. Den undersöker de unika och varierande kvalitéerna från den kontrasterande omgivningen. Tanken är att den nya stadshallen ska behålla den uppbrutna och intima känslan från den äldre bebyggelsen, och samtidigt bevara den solida känslan från stenstadskvarteret och domkyrkan. Resultatet är en uppdelad tegelbyggnad där omkringliggande mark binder samman de olika delarna. Relationen mellan de separata byggnadskropparna formar en serie av öppningar och utomhusytor. Dessa skapar i sin tur koppling, aktivitet och flöde mellan och runt om stadshallen.

  • Börgö, Arvid
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Civic Hall "Passagen"2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [sv]

    Strängnäs är en stad med rötter långt tillbaka i historien, och som har ett vackert och blandat stadslandskap. Uppgiften i mitt kandidatprojekt har varit att utforma ett förslag på medborgarhus, beläget på en tomtyta bredvid ett bankhus i kanten till Västervikstorget. Närheten till vattnet, rörelsen på de offentliga ytorna och småskalighetens relation till det mer storskaliga, är teman som har varit viktiga i utformandet av förslaget. Ambitionen med mitt förslag är att skapa en plats som både ger Strängnäs nya offentliga rum och funktionella ytor, men som även binder ihop den närmsta miljön och inbjuder till rörelse.

    Genom att arbeta med de siktlinjer och offentliga förbindelser som byggnadens tomrum skapar, väcker projektet frågor om hur vår mänskliga nyfikenhet och förväntan kan användas som arkitektoniskt grepp. Förslagets gestaltning tar också upp frågor om hur det intima kan möta det offentliga, hur öppenhet kan förenas med slutenhet.

  • Linde, Fredrika
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Kommunhuset - en plats för samling2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    The main purpose has been to create a welcoming, open and living municipal house next to Nynäshamn's main square. The building combines the two parallel uses - workplace and public space and allows these two to enrich each other. Visual connections between the floors in the building merge the public and private space together. 

  • Enegård, Lars-Erik
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Strengnes Stadshall2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [sv]

    Cirka 8 mil väster om Stockholm ligger Strängnäs. En mindre stad som har sina rötter i den svenska medeltiden. Då staden under många år inte haft någon egentlig publik samlingsbyggnad så blev idén med att uppföra en stadshall nere vid hamnområdet en positiv och spännande utmaning. De essentiella faktorerna som kom att specifikt ligga till grund för utformandet av "Strengnes stadshall" är därför följande:

    - Att utmana norrläget genom att låta ljuset/solen nå byggkropparnas "väsentliga" delar, samt omgivningen med boenden som idag har fullt exponerat söder- och västläge. Därför anpassas byggnadens höjd och takvinklar till solens bana.

    - Att skapa en fysiskt "lätt" och följsam struktur som samtidigt dockar minimalt mot den söderliggande tegelbyggnaden vilket samtidigt skapar en takgårdsplats som öppnar sig mot hamnområdet i väst och kvällssol.

    - Att arbeta efter en estetisk formel med en exteriör materialitet som allierar sig med den "lilla" staden och det torgliknande hamnområdet.

    Byggnaden består därav av två sammanlänkade byggkroppar med två huvudsakliga funktioner där begrepp som "nytta och nöje" samspelar i de olika rumsligheterna. Med hänsyn till fastighetens exponering mot bland annat den stora öppna torgytan åt väster har beaktande därför tagits i såväl det invändiga som det utvändiga rummet, både i termer av visuell och funktionell upplevelse.

  • Förell, Lycke
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Arkitektur.
    Fire Works2019Independent thesis Advanced level (degree of Master (One Year)), 10 poäng / 15 hpOppgave
    Abstract [en]

    Fire Works is a cultural center that will seem durable, welcoming and useful. The Culture center will serve as an extension of the central fireplace in the building. Something that gives light and life to the city with the activity created by the presence of the citizens. The building remains within the framework of the environment that surrounds it, a million program area with material dominance of concrete, glass and steel.

  • Lu, Yu-Chiao
    KTH, Skolan för industriell teknik och management (ITM), Materialvetenskap.
    Design of Bridgman unidirectional solidification furnace2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    The thesis work consists of two parts. First, the development of two-dimensional numerical models of a Bridgman unidirectional solidification furnace, and second, the construction work of the furnace at KTH. The aim is to build a Bridgman furnace which is capable of close control over temperature gradient and growth rate such that the solidification structures of a duplex stainless steel (SAF2507) could be replicated at a laboratory scale for different cooling rates.Two numerical models of Bridgman furnace are created using COMSOL Multiphysics. The models are used as predictive tools to simulate the locations of solidification front and the temperature gradients at the solidification fronts, which are parameters difficult to access during experiments. Different hot zone temperatures of the furnace (1500~1550 °C) and different sample pulling rates (0.5~10 mm/s) are studied in simulations. The major finding from modeled results is that the temperature gradient of the sample at the solidification fronts range from 5 ~ 17 K/mm, which are lower than the furnace temperature gradient of ~50 K/mm. The corresponding steady-state cooling rates range between 5 ~ 85 K/s. The next step is to validate the models with experimental temperature profiles of the furnace, and decide whether the furnace design should be modified to achieve the cooling rates of interests.

  • Stavrou, Fotios
    et al.
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Teknisk informationsvetenskap.
    Skoglund, Mikael
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Teknisk informationsvetenskap.
    Optimization and Tracking of Scalar-Valued LQG Control Under Communication Link with Synchronized or Delayed CaSI at the Decoder2019Annet (Annet vitenskapelig)
  • Söder, Isabelle
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Byggvetenskap, Transportplanering.
    Modelling Effects of Car Sharing on Travel Behaviour2019Independent thesis Advanced level (degree of Master (Two Years)), 20 poäng / 30 hpOppgave
    Abstract [en]

    Shared modes of transport, including car sharing, have been pointed out as one way of reducing private car use, contributing to an efficient transportation system that fulfills societal and environmental goals.Previous studies show that a share of car sharing users sells or refrains from acquire a new vehicle, when entering car sharing. Also, on average, car sharing has been shown to reduce Vehicle Kilometers Traveled (VKT) by car among the users.This study is conducted in three parts. First, a literature review of the effects of car sharing on travel behavior and car ownership is presented. Second, an implementation of car sharing in an existing transport model is described and the estimated effects are analyzed in relation to the findings in the literature study. In the final part, the car sharing module is reformulated to model a station-based car sharing system, where the distances to car sharing vehicles are used to distribute the effect of car sharing on car ownership spatially.This work contributes to the field by connecting the results from previous research about car sharing with practical transport modelling. The model of the station-based car sharing system is a useful tool for planners when considering the placement of car sharing stations. Also, this study provides an updated literature review covering findings of the effects of car sharing on travel behaviour and car ownership.Keywords: car sharing, station-based car sharing, travel demand modelling, vehicle ownership modelling, four-step model

  • Ceccato, Vania
    et al.
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Samhällsplanering och miljö, Urbana och regionala studier.
    Wilhelmsson, Mats
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Fastigheter och byggande, Bygg- och fastighetsekonomi.
    Do crime hot spots affect housing prices2019Inngår i: Nordic Journal of CriminologyArtikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Our knowledge about what happens to housing values when properties are close to places with high concentrations of crime, often called ‘hot spots’, is limited. Previous research suggests that crime depresses property prices overall, but crime hot spots affect house prices more than crime occurrence does and may affect prices of single-family houses more than prices of flats. Here we employ hedonic price modelling to estimate the impact of crime hot spots on housing sales, controlling for property, neighbourhood and city characteristics in the Stockholm metropolitan region, Sweden. Using a Geographic Information System (GIS), we combine property sales by coordinates into a single database with locations of crime hot spots. The overall effect on house prices of crime (measured as crime rates) is relatively small, but if its impact is measured by distance to a crime hot spot, the effect is non-negligible. By moving a house 1 km further away from a crime hot spot, its value increases by more than SEK 30,000 (about EUR 2,797). Vandalism is the type of crime that most affects prices for both multi- and single-family housing, but that effect decreases with distance from a crime hot spot.

  • Ceccato, Vania
    et al.
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Samhällsplanering och miljö.
    Vasquez, Lisandra
    Langefors, Linda
    KTH.
    Canabarro, Ana Paula
    Petersson, Robin
    KTH.
    Trygg stadsmiljö: Teori och praktik för brottsförebyggande & trygghetsskapande åtgärder2019Rapport (Annet vitenskapelig)
    Abstract [sv]

    I februari år 2019 erhöll Avdelningen för Samhällsplanering och Miljö på KTH i uppdrag av Boverket att kartlägga aktuell teori och praktik för brottsförebyggande och trygghetsskapande fysiska åtgärder i Sverige. Målet med studien är inhämta och sammanställa kunskap om hur trygghetsskapande och brottsförebyggande perspektiv ser ut idag och vilka åtgärder som vidtas i utformningen av fysiska miljöer i de olika skedena av samhällsbyggnadsprocessen.

    Den här rapporten bidrar till byggandet av kunskap inom hållbar stadsplanering med fokus på säkerhet och trygghet. Rapporten är en kartläggning av aktuell nationell och internationell teori och praktik för brottsförebyggande och trygghetsskapande fysiska åtgärder.

  • Ceccato, Vania
    et al.
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Samhällsplanering och miljö, Urbana och regionala studier.
    Näsman, Per
    KTH, Skolan för arkitektur och samhällsbyggnad (ABE), Fastigheter och byggande, Bygg- och fastighetsekonomi.
    Catherine, Sundling
    Langefors, Linda
    KTH.
    Trygghet i kollektivtrafiken i Stockholm i ett internationellt perspektiv: En handlingsplan mot sexuella trakasserier och brott i transitmiljöer2019Rapport (Annet vitenskapelig)
    Abstract [en]

    Safety and security are human rights - to feel free from risk and fear of danger is crucial for all people and a prerequisite for modern society. Safety promotes and encourages mobility, which is basic for an individual’s quality of life. This report presents Stockholm students' responses to a survey of experiences and perceptions of sexual harassment and sexual transit crimes in public transport conducted in May-June 2018 as part of an international study. The study contained 1,122 young people surveyed at the Royal Institute of Technology (KTH) in Stockholm and 309 students from Södertörn University in Huddinge. We investigated the victimization of sexual harassment and sexual crimes in transit environments, feelings of perceived security or lack thereof, and necessary improvements to make travel safer.

  • Kann, Viggo
    KTH, Skolan för elektroteknik och datavetenskap (EECS), Teoretisk datalogi, TCS.
    Programme Integrating Courses Making Engineering Students Reflect2019Inngår i: Theorizing STEM Education in the 21st Century, IntechOpen , 2019Kapittel i bok, del av antologi (Fagfellevurdert)
    Abstract [en]

    A programme integrating course (PIC) is a special type of course, lasting for several academic years and aiming to strengthen programme coherence, by tying the students, instructors and programme director closer together. The first PIC was started at KTH in 2008. Since then, the concept has been polished and adopted by many engineering and masters of science programmes at KTH and at other universities. The course is built around regular (four times a year) reflection seminars in small cross-grade groups, mentored by a teacher. Each seminar has a topic, for example, study skills, procrastination, exchange studies, generic skills, minorities and equal treatment and ergonomics and mental health. Before the seminar, the students are presented with some material to read and view. Based on the texts and videos, each student should write a reflection document and read and comment some other students’ reflections. At the seminar, the students will further discuss the topic and discuss the courses that they are currently taking. PIC has been evaluated and found very valuable by both the students and the teachers acting as mentors. This chapter will review the existing literature on PICs, which is mostly in Swedish.