Skip to content

README.md

Latest commit

 

History

History
146 lines (81 loc) · 3.52 KB

README.md

File metadata and controls

146 lines (81 loc) · 3.52 KB
Raw

Bitcoin Price Predictor 🔮

A machine learning web application that predicts Bitcoin price movements using Wikipedia sentiment analysis and technical indicators.

Features

  • Real-time Bitcoin price predictions
  • Wikipedia edit sentiment analysis
  • Automated daily model updates
  • Web interface for easy access
  • Confidence scoring

How It Works

  1. Analyzes Wikipedia Bitcoin page edits for sentiment
  2. Processes Bitcoin price data and technical indicators
  3. Uses XGBoost machine learning model
  4. Predicts next-day price movement (UP/DOWN)

Local Development

1. Clone the repository

git clone https://github.com/AlhassenSabeeh/bitcoin_predictor.git cd bitcoin_predictor

2. Create and activate virtual environment

python -m venv myenv source myenv/bin/activate

3. Install requirements

pip install -r requirements.txt

4. Update data and run the application

python update_data.py

after run python update_date.py, the following shoud be apearded in the terminal

🚀 BITCOIN PREDICTOR - DATA UPDATE TOOL

============================================================

🕐 Update started at: 2025-11-16 20:33:31

🔍 Checking system dependencies

✅ Basic dependencies check passed

==================================================

📊 WIKIPEDIA SENTIMENT DATA UPDATE

==================================================

No model was supplied, defaulted to distilbert/distilbert-base-uncased-finetuned-sst-2-english and revision 714eb0f

(https://huggingface.co/distilbert/distilbert-base-uncased-finetuned-sst-2-english).

Using a pipeline without specifying a model name and revision in production is not recommended.

Device set to use cpu

✅ Sentiment analysis pipeline loaded successfully

🚀 Starting Wikipedia sentiment analysis pipeline...

📥 Fetching Wikipedia Bitcoin page edits...

✅ Fetched 18069 Wikipedia revisions

🧠 Analyzing sentiment of Wikipedia edits...

📊 Found 16738 valid sentiment scores out of 18069 revisions

✅ Analyzed sentiment for 2799 days with 16738 valid scores

✅ Sentiment analysis complete. File saved as 'wikipedia_edits.csv'

✅ Wikipedia data update completed successfully

  • Data points: 5746
  • Date range: 2010-02-23 00:00:00 to 2025-11-16 00:00:00

==================================================

🤖 BITCOIN PREDICTION MODEL UPDATE

==================================================

🚀 Starting Bitcoin prediction pipeline...

📊 Loading Bitcoin price data from Yahoo Finance...

✅ Loaded 4079 days of Bitcoin data (up to 2025-11-16 00:00:00)

🔄 Merging price data with Wikipedia sentiment...

✅ Loaded sentiment data for 5746 days

✅ Successfully merged price and sentiment data

⚙️ Creating technical features...

✅ Created 15 features across 4 time horizons

🤖 Training XGBoost model...

📈 Training on 4079 samples with 15 features...

✅ Model training complete and saved

🎯 Prediction: DOWN (Confidence: 61.15%)

✅ Prediction pipeline complete!

✅ Bitcoin model update completed successfully

  • Features used: 15
  • Training date: 2025-11-16T20:55:50.637579

========================================

📋 UPDATE SUMMARY

========================================

🕐 Started: 2025-11-16 20:33:31

🕐 Finished: 2025-11-16 20:55:50

⏱️ Duration: 1338.87 seconds

🎯 PREDICTION RESULTS:

Next day prediction: DOWN

Confidence: 61.15%

Current Price: $94,127.77

UP Probability: 38.849998474121094%

DOWN Probability: 61.150001525878906%

Model Training: 2025-11-16

✅ Update process completed!