Solution to MATLAB and Simulink Challenge project

March 6, 2025 · View on GitHub

https://github.com/steven1he/BTC-price-prediction-using-sentimental-analysis

Program details

Overview

This project aims to predict Bitcoin (BTC) prices using sentiment analysis of cryptocurrency-related text data, combined with historical price data. The workflow includes data preprocessing, feature engineering, model training, and evaluation.

Demo

result picture

Workflow, Directory Structure, Requirements, Usage, and Results


1. Download Datasets:
   - BTC Tweets Dataset: https://www.kaggle.com/datasets/kaushiksuresh147/bitcoin-tweets
   - BTC Historical Price Data: https://www.kaggle.com/datasets/mczielinski/bitcoin-historical-data#coinbaseUSD_1-min_data_2014-12-01_to_2019-01-09.csv
   - You can simply use the outcome 'filtered_merged_sentiment_price1.csv' for trianing and testing.

2. Run Sentiment Analysis:
   - Execute the script sentimental_score.m to calculate sentiment scores (e.g., maximum, average, variance) for all tweets over the past 90 days, aggregated by minute.

3. Preprocess Historical Data:
   - Use the script shift_time.m to convert timestamps into a human-readable format (year, month, day, hour, minute).

4. Merge Data:
   - Run merge_data.m to combine the historical price data and the sentiment analysis results into a single dataset: merged_file_90days_1min(final).csv.
   - Note: The merged file is already included in the repository for direct use.

5. Train and Test the Model:
   - Run train_test.m to train the prediction model and evaluate its performance.
   - Prediction results are saved as plots in the result_picture/ directory.

Directory Structure:
BTC-price-prediction-using-sentimental-analysis/
├── sentimental_score.m        # Sentiment score calculation script
├── shift_time.m               # Timestamp conversion script
├── merge_data.m               # Data merging script
├── train_test.m               # Model training and testing script
├── merged_file_90days_1min(final).csv  # Processed dataset for training
├── result_picture/            # Directory containing prediction results

Requirements:
- MATLAB R2021a or later
- Required MATLAB Toolboxes:
  - Statistics and Machine Learning Toolbox
  - Deep Learning Toolbox

Usage:
1. Clone the repository:
   git clone https://github.com/steven1he/BTC-price-prediction-using-sentimental-analysis.git
   cd BTC-price-prediction-using-sentimental-analysis

2. Run the following MATLAB scripts in sequence:
   matlab sentimental_score.m
   matlab shift_time.m
   matlab merge_data.m
   matlab train_test.m

Results:
- The final processed dataset: merged_file_90days_1min(final).csv
- Prediction results, including plots of actual vs. predicted prices, are saved in the result_picture/ directory.