Back to Library
Naive Bayes Trend Classifier
AI / MLUses Bayesian probability to classify the current market state as Bullish, Bearish, or Neutral by combining evidence from momentum, volume, and volatility features.
How It Works
The indicator computes likelihoods for three market classes (Bull, Bear, Neutral) based on independent features: RSI zone, volume relative to average, ATR expansion/contraction, and close position within the recent range. Using Bayes' theorem, it multiplies prior probabilities by each feature's likelihood to produce posterior probabilities for each class. The class with the highest posterior wins. A confidence gauge shows how dominant the winning class is, filtering out low-conviction signals.