Going Deeper with Convolutional Neural Network for Stock Market Prediction

Repository for Going Deeper with Convolutional Neural Network for Stock Market Prediction

Introduction

Predict the stock market price will go up or not in the near future.

Data Collection

• Using Yahoo! Finance for time series data source
• 50 Taiwan Companies from 0050.TW index.
• Top 10 Indonesia Stock exchange companies.

Methodology

• Using candlestick chart for input model
• DeepCNN
• ResNet 50
• VGG16
• VGG19
• Randomforest
• KNN

Usage

Prepare Environment

Recommended using virtual environment

python3 -m venv .env

Running on Python3.5

pip install -U -r requirements.txt

Prepare Dataset

• Convert OHLCV stock market data to Candlestickchart

  python run_binary_preprocessing.py <ticker> <tradingdays> <windows>

  example

  python run_binary_preprocessing.py 2880.TW 20 50

• Generate dataset

  python generatedata.py <pathdir> <origindir> <destinationdir>

  example

  python generatedata.py dataset 20_50/2880.TW dataset_2880TW_20_50

• Remove alpha channel

  cd /dataset/dataset_2880TW_20_50
  find . -name "*.png" -exec convert "{}" -alpha off "{}" \;

Training

• DeepCNN

  python myDeepCNN.py -i <datasetdir> -e <numberofepoch> -d <dimensionsize> -b <batchsize> -o <outputresultreport>

  example

  python myDeepCNN.py -i dataset/dataset_2880TW_20_50 -e 50 -d 50 -b 8 -o outputresult.txt

Performance Evaluation

• Accuracy
• Specitivity
• Sensitivity
• MCC
• F1

Citation

@misc{1903.12258,
Author = {Rosdyana Mangir Irawan Kusuma and Trang-Thi Ho and Wei-Chun Kao and Yu-Yen Ou and Kai-Lung Hua},
Title = {Using Deep Learning Neural Networks and Candlestick Chart Representation to Predict Stock Market},
Year = {2019},
Eprint = {arXiv:1903.12258},
}