Neural-Structure Momentum Hybrid

Family: hybrid · Regime: trending · Complexity: high · Asset classes: FX, US Equities (Large Cap), Crypto (Major Pairs) · Timeframes: 15m, 1H

Thesis

Market trends are most 'efficient' when a structural pivot is accompanied by a tick-level momentum burst. By utilizing KNN and LSTM models to filter out low-quality volatility and noise, we can identify entries where institutional participation is highest. The use of Darvas Boxes provides an objective, non-arbitrary risk level that adapts to market structure rather than fixed pip distances.

Components

• Volume SuperTrend AI (Expo) (regime) — Uses KNN classification to ensure the market is in a high-probability trending regime based on historical proximity to VWMA-based baselines.
• MultiKAMA (TyphooN) (direction) — Provides a volatility-adjusted directional bias; the Efficiency Ratio (ER) ensures we only trade when price movement is 'efficient' relative to noise.
• Bullish Pivot Detector (entry) — Identifies structural market turns where local resistance is breached after a specific candle sequence, providing the final execution trigger.
• Relative Strength Quality Index (RSX) (exit) — The Jurik-smoothed oscillator allows for noise-free exits when momentum reaches exhaustion (OS/OB) or rolls over.
• Darvas Boxes (Multi-Mode) (risk) — Establishes the risk boundary. The box floor (for longs) or ceiling (for shorts) provides a structural anchor for stop-loss placement.
• Flexible Momentum (confirmation) — Confirms the entry with a sub-minute 'burst' of tick-level momentum, reducing the risk of entering late into a fading candle.
• ML SuperTrend (Ultimate) - Auto-Optimized AI with LSTM (volatility_filter) — Functions as a dynamic volatility filter, suppressing entries if the LSTM/DQN model determines trend quality is decaying or volatility is idiosyncratic.

Known failure conditions

• Sustained decrease in tick-level volatility rendering the Flexible Momentum threshold unreachable.
• Model drift in the LSTM volatility filter causing it to stay in a permanent 'off' state during trending periods.
• Whipsaw price action that stays within Darvas Boxes but repeatedly triggers and stops out the Pivot Detector signals.

Explore the full interactive blueprint, parameter ranges and evidence on WOBR StrategyVerse, or generate this strategy as an MT4/MT5 Expert Advisor with QuantMogul AI Engine.