Institutional Repository of Vilnius University
Mašininio mokymosi panaudojimas akcijų kainų prognozavimui /
Abstract
dc:descriptionThis bachelor thesis shows that technical analysis and machine learning models can be used to predict future stock prices. Accurately predicting stock prices is very difficult but it could be highly profitable. Four technical analysis models were used: relative strength index, stochastic oscillator, candlestick models and moving average. Three machine learning models were used in this thesis: artificial neural network, support vector machine and random forest. Machine learning models were made with “WEKA“ toolkit. Stock prices were used as the research object of this work. For the technical analysis one year historical S&P 500 prices were used, for machine learning models 2007 days of historical data and technical analysis results were used. The theory regarding technical analysis and machine learning are reviewed. Best results of technical analysis were shown by relative strength index with 66.67% correct buy/sell signals. Best results of machine learning models were shown by support vector machine it gave the best accuracy results measured by RMSE and MAPE indicators compared with two other machine learning models. Machine learning models performance depends on the input data, model parameters, complexity. To predict prices and start trading one should combine technical analysis, machine learning and fundamental analysis to make profit.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Petruškevičius, Simas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD193299542&prefLang=en_US