Application of machine learning technique versus stastistical technique in financial forecasting: an empirical evidence from bursa Malaysia KLCITools Loh, Xuejing (2020) Application of machine learning technique versus stastistical technique in financial forecasting: an empirical evidence from bursa Malaysia KLCI. [Dissertation (University of Nottingham only)]
AbstractThe FTSE Bursa Malaysia Kuala Lumpur Composite Index (KLCI) is a market-capitalisation weighted index that consists of 30 largest listed companies. The KLCI is commonly used as the benchmark for the general market performance of Malaysian stocks. In spite the claims that the financial market is an efficient market and that it follows a random walk process and as a result, the stock prices are not predictable, there are literatures that claimed that the financial market can be predicted. In this management project study, it examines and assesses on the forecasting ability of statistical technique using the Autoregressive Integrated Moving Average (ARIMA) model against the machine learning technique using the Artificial Neural Network (ANN) model to predict the price movement of KLCI. The dataset comprises the historical Opening, High, Low and Closing (OHLC) prices of KLCI spanning from Jan 4, 2016 to Dec 31, 2018. The data collected is being categorised into two separate parts, training and testing sets; 90% of the data set makes up the training set while the remaining 10% data was used as testing sets. The experiment was carried out using IBM SPSS statistics V25.0 for modelling the ARIMA model and Jupyter Notebook V5.7.8 for modelling the ANN model. The research result shows that the ANN (7-5-4-1) outperformed the ARIMA (2,1,2) with higher accuracy level at 65% compared to 43%. Hence, the research concludes that the ANN model performed better than the ARIMA model in forecasting the KLCI in the studied context. This provides investors the viability to use the ANN model as an alternative to the ARIMA model in forecasting the KLCI.
Actions (Archive Staff Only)
|
Tools
Tools
![[thumbnail of Management Project - 20129537.pdf]](https://eprints.nottingham.ac.uk/style/images/fileicons/application_pdf.png "Management Project - 20129537.pdf")
