Neural-Darvas LVN Breakout Ensemble

Family: breakout · Regime: trending · Complexity: high · Asset classes: Crypto (ATOMUSDT focus), Equities, FX · Timeframes: 15m (Optimized for ATOM), 1H

Thesis

Breakouts from consolidation zones (Darvas) are statistically more likely to transition into sustained trends when they occur at high-volume nodes (Supply/Demand) and are confirmed by both linear (Logit) and non-linear (Neural Network) probability models, provided the market's forecast error (MdAE) remains low. The edge exists in the convergence of structural breakout, liquidity validation, and machine-learned momentum.

Components

• Total Power Indicator (regime) — Acts as the primary regime filter to ensure trades are only taken when a trend is establishing (Total Power > 20) and not in a 'dead' or perfectly balanced market.
• Darvas Boxes Modern/Classic (direction) — Identifies structural breakouts from consolidation ranges; the 'modern' mode allows for more adaptive box tracking.
• Dynamic Supply and Demand Zones [AlgoAlpha] (entry) — Filters breakouts to ensure price is interacting with Low Volume Nodes (LVN) or supply/demand shifts, providing a 'liquidity' anchor for the entry.
• Relative Vigor Index (RVI) (exit) — Used for momentum exhaustion; a crossover against the trend suggests the 'vigor' of the move is fading.
• ATOM Coefficient Indicator (logit) (risk) — Provides the final statistical probability layer and defines the risk parameters (SL/TP) using its hardcoded ATR coefficients.
• Neural Network Indicator (NNind) (confirmation) — Acts as a secondary confirmation filter, ensuring price action momentum (via CLV) is aligned with a pre-trained MLP model.
• Median Absolute Error (volatility_filter) — Filters out periods where the market is behaving erratically (high forecast error between Open and Close), ensuring higher predictability.

Known failure conditions

• The ATOM Coefficient logit model P-value remains between 0.45 and 0.55 for extended periods, resulting in zero signals.
• Market volatility (MdAE) stays consistently above its 50-period average, indicating a permanent shift to a high-noise regime where breakouts are statistically random.
• Consecutive Darvas Box collapses where price breaks one side and immediately reverses to the other (choppiness).

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