๐ BNB ML Features Dataset (2019-2025) 255,121 rows of processed BNB market data. ๐ Contents: 6 timeframes: 5m to 1d 19 engineered features Exchange token dynamics Regime detection labels ๐ฏ Use cases: BNB-specific trading models Binance ecosystem analysis Stablecoin pair research Largest single-asset dataset in this collection.
๐ Solana ML Features Dataset (2019-2025) 245,003 rows of processed SOL market data. ๐ Contents: 7 timeframes: 5m to 1d 19 features including tail_lambda High-volatility regime labels Momentum and mean-reversion signals ๐ฏ Use cases: Altcoin-specific models Cross-asset correlation studies Scalping strategy development Covers SOL's full price history.
๐ Ethereum ML Features Dataset (2019-2025) 245,003 rows of processed ETH market data with ML-ready features. ๐ Contents: 7 timeframes: 5m, 15m, 30m, 1h, 4h, 12h, 1d 19 engineered features Volatility expansion signals Drawdown and momentum indicators ๐ฏ Use cases: DeFi strategy optimization ETH/BTC correlation analysis Risk-adjusted position sizing Full 5-year history. Clean, labeled data.
๐ Bitcoin ML Features Dataset (2019-2025) 245,003 rows of processed Bitcoin market data with ML-ready features. ๐ Contents: 7 timeframes: 5m, 15m, 30m, 1h, 4h, 12h, 1d 19 engineered features Hindsight oracle labels (optimal_exposure) Tail risk indicators, volatility ratios, regime detection ๐ฏ Use cases: Train exposure prediction models Backtest BTC trading strategies Volatility regime classification Full 5-year history. Production-ready format.
๐ Fresh data from today's market (Dec 20, 2025) Real-time proof that the Antifragile pipeline is active. ๐ Contents: BTC, ETH, SOL (1h timeframe) Last 24 hours of processed features Same methodology as full dataset ๐ Updated: December 20, 2025 This is a live demonstration of our data processing system. For historical backtesting, see the full 5-year dataset.
๐งช Free sample from the Antifragile ML Features Dataset 10,000 rows randomly sampled from 990,130 total records. ๐ Contents: 4 assets: BTC, ETH, SOL, BNB Timeframes: 5m to 1d 19 engineered features Hindsight oracle labels (optimal_exposure) Date range: 2019-2025 ๐ฏ Use cases: Train XGBoost/LightGBM models Validate feature engineering pipeline Test your risk management strategies ๐ Full dataset available: โ 990,130 rows | 5 years of data | $49 Built for quants. Verified by backtesting.
990K premium labeled trading bars for BTC/ETH/SOL/BNB (2019-2025). Includes validated Machine Learning features: volatility regimes, tail risk lambda, and hindsight-optimal exposure. Verified IC Score: 0.775. Perfect for ML/RL bot development.
No organizations yet