Artificial intelligence in financial time series forecasting - A quantitative forecast of the OMXH25 index
No Thumbnail Available
URL
Journal Title
Journal ISSN
Volume Title
School of Business |
Bachelor's thesis
Electronic archive copy is available locally at the Harald Herlin Learning Centre. The staff of Aalto University has access to the electronic bachelor's theses by logging into Aaltodoc with their personal Aalto user ID. Read more about the availability of the bachelor's theses.
Authors
Date
2019
Department
Major/Subject
Mcode
Degree programme
(Mikkeli) Bachelor’s Program in International Business
Language
en
Pages
42+2
Series
Abstract
Objectives The main objective of this study was to analyze and evaluate the effectiveness of artificial intelligence applications in financial services. The scope of the research is to evaluate the applications of AI tools specifically in Finnish markets, to demonstrate adaptability into various market conditions, and to fill a geographic literature gap in financial AI applications. Summary An extensive literature review analyzes recent major publications in the field and builds a conceptual framework based on findings. Common methodology in building and evaluating intelligent computational tools is used, during which 27 input variables are chosen to forecast the OMXH25 stock market index. A NARX neural network model is chosen and trained with 250 time-steps of data for evaluation. Conclusions Findings suggest that intelligent AI tools can be easily built and adapted to various market conditions. An appropriate MSE of 0.023607 is obtained when modelling the OMXH25 index output values. The black-box neural network model provides limited insight into market structures and input variable weights but demonstrates the deteriorating accuracy when forecasting longer time-spans.Description
Thesis advisor
Grinsted, SusanKeywords
artificial intelligence, machine learning, financial forecasting, neural networks