parthpromax / stock-market-price-prediction

Predication of stock market price using different machine learning models

The model is done as follows.

1. ARIMA model

   i) First run preprocess.py with command : python preprocess.py This will create 5 files which correspond to different data. Original,Log,Moving Averge,Exponential Average & Decomposed For the convenience we have already included all the files already in the directory.

   ii) Now run once - arma.R and arima.R one by one. which will calculate the optimum (p,i,q) value in case of arima and (p,0,q) in case of arma.R. These files will also predict the share price on the next 10-days. Again for the convenience, all the codes which produce and save graphs have been either commented or made manually.

2. ANN

   i) First run Preprocessing_of_data.py with the command : python Preprocessing_of_data.py This file will generate all the technical parameters like RollingAvg, MovingAvg etc. in the file train.csv ii) ANN.py now will take train.csv as a input and train the neural network on parameters to produce the accuracy.

3. SVC

   i) First run Preprocessing_of_data.py with the command : python Preprocessing_of_data.py This file will generate all the technical parameters like RollingAvg, MovingAvg etc. in the file train.csv ii) SVC.py will now take train.csv as a input and train the Support vector classifier to produce accuracy on given data.

For more details, you can go through report and presentation.