In English

Stock Market prediction using Artificial Neural Networks

Rafael Konstantinou
Göteborg : Chalmers tekniska högskola, 2018. 83 s.
[Examensarbete på avancerad nivå]

Stock market is one of the most competitive financial markets and traders need to compute the financial workloads with low latency and high throughput. In the past, people were using the traditional store and process method to calculate the heavy financial workloads efficiently. However to achieve low latency and high throughput, data-centers were forced to be physically located close to the data sources, instead of other more economically beneficial locations. This is the main reason, the data-streaming model was developed and it can process large amount of data more efficiently. It was shown in studies that using data streaming we can solve the options pricing and risk assessment problems using traditional methods, for example Japanese candlesticks, Monte-Carlo models, Binomial models, with low latency and high throughput. However instead of using those traditional methods, we approached the problems using machine learning techniques. We tried to revolutionize the way people address data processing problems in stock market by predicting the behaviour of the stocks. In fact, if we can predict how the stock will behave in the short-term future we can queue up our transactions earlier and be faster than everyone else. In theory, this allows us to maximize our profit without having the need to be physically located close to the data sources. We examined three main models. Firstly we used a complete random model using a monkey trader. The name monkey trader comes from B.G. Malkiel’s claim, that a blindfolded monkey throwing darts at a newspaper’s financial pages could select a portfolio that would do just as well as one carefully selected by experts. It works by choosing random actions at random moments. Secondly we used a feed-forward artificial neural network (ANN) model and finally a model that uses Reinforcement Learning(RL). Each of those models was applied on real stock market data and checked whether it could return profit.

Nyckelord: Stock market, Artificial Neural Networks, Machine Learning



Publikationen registrerades 2018-10-10. Den ändrades senast 2018-10-10

CPL ID: 256121

Detta är en tjänst från Chalmers bibliotek