ML SuperTrend (Ultimate) - Auto-Optimized AI with LSTM

Category: trend, volatility

An advanced SuperTrend implementation that utilizes a Deep Q-Learning (DQN) architecture with LSTM layers and K-Means++ clustering to dynamically optimize parameters based on volatility and trend quality.

Formula

SuperTrend = (Source \pm (ATR \times Factor)); \text{Factor} = f(LSTM(Features), K\text{-Means}(Clusters)) \text{ where Features include ROC, ATR, and Normalised price data.}

Inputs

• resample_freq (default: 3)
• atr_length (default: 10)
• base_fact (default: 1.3)
• lstm_hidden_size (default: 8)
• learning_rate (default: 0.01)

See signal primitives, usage in published strategies and more on WOBR StrategyVerse.