Skip to content

SC1015-A139-TEAM7/mini-project

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 
 
 
 
 
 
 

Repository files navigation

SC1015 Mini Project

About

This is a Mini-Project for SC1015 (Introduction to Data Science and Artificial Intelligence) which has the objective of finding the best methodology to predict stock price.

We used the S&P 500 stock data taken from Alphavantage API. Data extraction and preprocessing is in this file: link the data preprocessing notebook

You can find the video explanation of our mini project here

Contributors

  • @joankadevi - Data Visualisation, Data Analysis, MA, LSTM, Slides and Video editing
  • @JeremyNathanJusuf - Data Extraction, NAIVE, MA, Exponential Smoothing, LSTM, Data Analysis

Problem Definition

In today's fast-paced world, the stock market is one of the most popular avenues for investments. However, predicting stock prices accurately is a challenging task that requires advanced techniques and tools as the stock market is complex and volatile.

Investors are looking for an accurate and reliable predictive model which can capture the market’s highly volatile and nonlinear behavior.

That's where our project comes in. Our objective is to find the best forecaster to predict stock prices using various techniques and a machine learning model. By doing so, we hope to help individual investors make better-informed decisions about their investments.

Statistical and Machine Learning Models

  1. NAIVE METHOD
  2. MOVING AVERAGE
  3. EXPONENTIAL SMOOTHING
  4. LONG SHORT-TERM MEMORY NETWORKS

Conclusion

We used RMSE as the model performance metric for our experiment. RMSE values obtained by our models (in USD):

  1. NAIVE METHOD: 4.617
  2. MOVING AVERAGE WINDOW SIZE 5: 6.472
  3. EXPONENTIAL SMOOTHING WITHOUT TREND: 4.599
  4. EXPONENTIAL SMOOTHING WITH TREND: 4.631
  5. LSTM: 7.299

Hence, we can conclude that the best method to predict future stock price is exponential smoothing without trend as it has the lowest RMSE value indicating that it is the most accurate method. However, this such RMSE is still considered large in the context of predicting stock prices and would need further improvements to produce more accurate predictions.

Further Improvements

The LSTM model is a very promising model to predict stock price as well. In this case we can actually improve our model of LSTM by implementing other learnable indicators for the model to study from, for example the relative strength index, USD index, news, consumer sentiment index, unemployment rate, etc. Besides, we can tune better hyper parameters for the LSTM model to be more accurate.

What we learnt from this project

  • Theory of stock movements
  • Collecting data through API request
  • Recurrent Neural Networks (Particularly LSTM), Keras, Tensorflow
  • Mathematical conceptsof NAIVE, Moving Average, Exponential Smoothing and LSTM
  • Hyperparameters of models mentioned and how to find the most optimal one
  • Implementation of the statistical and ML models mentioned in Python
  • Project collaboration using GitHub

References

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published