MW192037
Stock Prediction

The stock prediction helps investors in making investment decisions. The model implemented is LSTM which is a feed-forward neural network. There are four main features in the system. The first feature is chart analysis to let investors analyze the stock through the candlestick graph. Also, the moving average of the stock can be plotted with the line graph. Next, the prediction function is to predict the stock based on the stock and date selected by investors. The validation function is a function to display the performance of the model by showing the predicted result within a range of period selected. Lastly, customization functions to let investors customize their model with re-train the existing model and load it to predict the stock. Since, the model is trained based on the sector of the stock, investors can focus on a certain stock that they interest in to customize their own model.

Name: Chai Ai Fang

Matrics No: 132885

Email: aifangchai_0711@student.usm.my

​

Supervisor: Dr. Chan Huah Yong

Email: hychan@usm.my