Method of and apparatus for automated behavior prediction
View Patent ↗A computer-implemented method of behavior prediction includes selecting behavior examples having corresponding antecedent candidates, identifying source text descriptions describing the behavior examples, automatically extracting predictors as common themes across all statements and all behavior examples with a language-independent theme extraction process, flagging each behavior example to indicate a presence or absence of the corresponding extracted antecedents in each of the source text descriptions and creating a data array consisting of antecedent columns and behavior example rows, submitting the data array to a pattern classifier to extract patterns among the antecedent candidates and outcomes by training and validating the pattern classifier and predicting a new occurrence of a target behavior by entering a current state of the antecedents to the trained pattern classifier.
1. A computer-implemented method of behavior prediction, comprising:
selecting behavior examples having corresponding precursor antecedent candidates;
identifying source text descriptions describing the behavior examples;
associating text statements from the source text descriptions for each behavior example with each of its corresponding precursor antecedent candidates;
flagging each behavior example for each associated text statement to indicate a presence or absence of the corresponding precursor antecedent candidates in each of the source text descriptions;
training a pattern classifier to extract patterns among the precursor antecedent candidates constituting inputs to the pattern classifier and outcomes constituting outputs of the pattern classifier to produce weights associating the inputs with the outputs;
optimizing and validating different pattern classifiers by using different combinations of antecedents as inputs, different pattern classifier configurations, and determining a highest performing pattern classifier as determined by accuracy metrics; and
predicting a new occurrence of a target behavior by inputting to the highest performing pattern classifier a vector describing a current state of antecedents for the highest performing pattern classifier, or different combinations of antecedents for the highest performing pattern classifier in a batch mode.
2. The computer-implemented method of claim 1 , further comprising summarizing the associated text statements and the corresponding antecedent candidates to identify themes across the behavior examples.
3. The computer-implemented method of claim 2 , wherein the summarizing is language-independent.