Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Forecasting Shanghai Composite Index using hidden Markov model
KTH, School of Engineering Sciences (SCI), Mathematics (Dept.), Mathematical Statistics.
2017 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Analys av Shanghai Composite Index med hjälp av dolda Markov modeller (Swedish)
Abstract [en]

The aim of this thesis is to forecast the future trend of Shanghai Composite Index and other securities. Our approach is applying the hidden Markov models to the market-transaction data indirectly. Previous work has not consider the independent problem between each training samples, which may result in inference bias. So we should select samples which is not significantly dependent, and suppose those samples are independent to each other. Rather than forecasting the future trend by estimating the hidden state one day before the trend, we measure the probabilities of the trend directions by calculating the gaps between the likelihoods of two hidden Markov models in a periods of time before the trends. As we have altered the target function of the optimization in parameter-estimation process, the accuracy of our model is improved. Furthermore, the experiment result reveals that it is lucrative to select securities for portfolios by our method.

Abstract [sv]

Syftet med detta arbete är att prediktera den framtida utvecklingen av Shanghai Composite Indexoch andra värdepapper. Vår strategi är att tillämpa de dolda Markov modeller tillmarknadstransaktionsdata indirekt. Tidigare arbete har inte överväga oberoendeproblem mellan varje träningsprov, vilket kan resultera till inferens-bias. så vibör välja datapunkter som inte är signifikant beroende med varandra. Istället för att prediktera framtida utvecklingen genom att skatta det dolda tillståndet en dag före trenden, mäter vi sannolikheterna för trendensriktningar genom att beräkna gapet mellan sannolikheterna för två dolda Markov Modelleri en tid före trenderna. Eftersom vi har förändrat målfunktionen avoptimering i parameter skattningsprocessen har riktigheten i vår modell förbättrats.

Place, publisher, year, edition, pages
2017.
Series
TRITA-MAT-E, 2017:07
National Category
Probability Theory and Statistics
Identifiers
URN: urn:nbn:se:kth:diva-204021OAI: oai:DiVA.org:kth-204021DiVA: diva2:1083813
External cooperation
Valley Oak
Subject / course
Mathematical Statistics
Educational program
Master of Science - Applied and Computational Mathematics
Supervisors
Examiners
Available from: 2017-03-22 Created: 2017-03-22 Last updated: 2017-03-22Bibliographically approved

Open Access in DiVA

No full text

By organisation
Mathematical Statistics
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar

Total: 63 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf