BitNexus - AI Bitcoin Price Prediction Platform
A Bitcoin price prediction research platform combining multi-modal deep learning (LSTM, BiLSTM, GRU, BiGRU) with sentiment fusion - a forecasting research pipeline with a simulated trading dashboard.
Key Outcome
5–10% RMSE reduction
Fusion models outperform price-only baselines with 3+ point directional accuracy gain.
Existing Bitcoin prediction tools rely solely on price and volume data, ignoring the strong market-moving influence of social media sentiment and news. Traders lacked a unified platform for AI-driven forecasts with a simulated trade view.
Built BitFusion, a research pipeline fusing OHLCV technical indicators with Twitter/news sentiment (VADER, TextBlob, AFINN, NRC) into multi-modal LSTM/BiLSTM/GRU/BiGRU architectures - then deployed the best model into BitNexus, a research prototype with a live prediction dashboard and simulated trading engine.
Key Features
Multi-modal fusion: OHLCV + sentiment (VADER, TextBlob, AFINN, NRC)
LSTM, BiLSTM, GRU, BiGRU and hybrid MFB classifier architectures
90-run factorial experiment sweep with tracked model registry
Live 1-hour Bitcoin price forecasts with confidence scoring
Simulated trading mode with configurable stop-loss and take-profit
Real-time BTC/USDT candlestick chart with TradingView integration
JWT-authenticated user dashboard with model selection
Hugging Face Space deployment with predict.py inference API
Technology Stack
Project Visuals
Interested in a similar project?
Tell us your requirements and we will respond within one business day.