Intelligent Systems: Report

by asokawotulo on 22/01/2020

Introduction

For our final project our team consisting of Asoka, Arnold & Nixon decided to create a program which predicts how the stock market would move. We decided to create after throwing away many ideas. We coded this using python, matplotlib, numpy, pandas, scipy, dotenv, keras, tensorflow, and alpha-vantage.

Weekly Report

On the first week of the announcement of the final project we planned and thought about a topic/program we wanted to create for the project. Initially, we brainstormed the idea and had options on stock market prediction, image recognition, etc. But we decided on stock market prediction because it has more value to people of our age and we feel like we could use it after the stock market.

The following week after we started to work on our final project researching how the stock market works and what we needed to successfully make the program run. We researched on what algorithms best suit stock market algorithm.  

And on the third week we started the coding process on creating this program, we decided on using LSTM algorithm to predict the stock market. We also needed to take all the data of the stock market so we used an online API called AlphaVantage. The data provided in the API has a few classifications and we chose to use intraday data which is the data per minute in the stock market for more detail.

For the fourth week and so on we tried to improve our program giving it some more data to make it more accurate, we also added in some technical indicators such as RSI, MACD and Bollinger Bands to make it more accurate.

Conclusion

The program we created was intended for personal use, and the results were not always good on every stock symbol but we tried to feed it more data to further improve the results.

asoka.watulo@binus.ac.id

Comments are closed.