Autonomous-Stock-Trading-Using-LSTM-Models

This repository implements an autonomous stock trading application that uses Long Short-Term Memory (LSTM) neural network models to make stock price predictions. Note that the code and results in this repository are for educational and experimental purposes only; always conduct your own research and tests before making real financial decisions.

Table of Contents

1. Overview
2. Requirements
3. Results
   • Model Evaluation Results
   • Predictions
   • Trading Results
4. Acknowledgements

Overview

This project aims to:

• Use LSTM-based deep learning models to predict price movements.
• Automate buy/sell decisions based on predicted trends. The main goal is to explore the use of deep learning in trading strategies.

For more details you can check the project report

Requirements

You can download the necessary libraries from the requirements.txt with this command: pip install -r requirements.txt.

Results

Model evaluation results

Stock Name	MAE	MSE	RMSE
ASELS	0.3169	0.2575	0.5074
THYAO	0.3064	0.2622	0.5120
AEFES	0.4566	0.5370	0.7328
AFYON	0.1074	0.0239	0.1547

---

Predictions

Prediction for AEFES

Prediction for AFYON

Prediction for ASELS

Prediction for THYAO

---

Trading Results

Trading of ASELS

Trading of THYAO

Daily portfolio value while trading of ASELS and THYAO

Acknowledgements

• Dataset: https://www.kaggle.com/datasets/gokhankesler/borsa-istanbul-turkish-stock-exchange-dataset/data