Reversals
The Reversal Engine is one of the most advanced components of Momentum Concepts. It uses a K-Nearest Neighbors (KNN) algorithm to calculate the statistical probability of a top or bottom.
How KNN Works
Feature Extraction: It tracks 8 dimensions of momentum (CCI over multiple timeframes).
Historical Comparison: Every new candle, the engine looks back at your "Max Samples" and finds the "K" most similar historical patterns.
Probability Scoring:
If most similar patterns in the past led to a bottom, the Bottom Probability increases.
If they led to a top, the Top Probability increases.
Signal Threshold
By default, the Signal Threshold is set to 80%. When the probability of a reversal crosses this line, the engine is "Primed".
Reversal Markers
Upward Triangle: Triggers when the Bottom Probability crosses back under the threshold after reaching an extreme. Indicates a confirmed bottoming pattern.
Downward Triangle: Triggers when the Top Probability crosses back under the threshold after reaching an extreme. Indicates a confirmed topping pattern.
Sensitivity
K-Value: Lower values make the engine more sensitive; higher values make it more stable.
Max Samples: Increasing this provides more historical context but requires more processing power.
Last updated
Was this helpful?