def _importance(self, feature): prior_prediction = self.predict( [util.bias_feature()] + [pb.Feature()] * (self._config.num_features - 1)) with_weight_prediction = self.predict( [util.bias_feature()] + [pb.Feature()] * (self._config.num_features - 2) + [feature]) return util.kl_divergence(with_weight_prediction, prior_prediction)
def _importance(self, feature): prior_prediction = self.predict([util.bias_feature()] + [pb.Feature()] * (self._config.num_features - 1)) with_weight_prediction = self.predict([util.bias_feature()] + [pb.Feature()] * (self._config.num_features - 2) + [feature]) return util.kl_divergence(with_weight_prediction, prior_prediction)
def __init__(self, config): self._config = config self._weights = {} # Initial bias weight bias_weight = util.prior_bias_weight(config.prior_probability, config.beta, config.num_features) self._set_weight(util.bias_feature(), bias_weight)
def __init__(self, config): self._config = config self._weights = {} # Initial bias weight bias_weight = util.prior_bias_weight( config.prior_probability, config.beta, config.num_features) self._set_weight(util.bias_feature(), bias_weight)
def _create_feature_vector(num_features): return [util.bias_feature()] + \ [pb.Feature()] * (num_features - 1)