The Theory of Probabilistic Hierarchical Learning for Classification

Providing the ability of classification to computers has remained at core faculty artificial intelligence. Its application now made inroads towards nearly every walk life, spreading over healthcare, education, defence, economics, linguistics, sociology, literature, transportation, agriculture, and industry etc. To our understanding most problems faced by us can be formulated as problems. Therefore, any novel contribution in this area a great potential applications real world. This paper proposes way learning from datasets i.e., hierarchical through set partitioning. The theory probabilistic for been evolved several works while widening its scope with each instance. demonstrates that dataset learnt generating hierarchy models capable classifying disjoint subset training set. basic assertion behind is an accurate complex achieved low complexity models. In paper, redefined revised based on four mathematical principles namely, principle successive bifurcation, two-tier discrimination, class membership selective data normalization. algorithmic implementation also discussed. approach further widened include ten popular real-world test base. does not only produce their but produced above 95% accuracy average regard generalising ability, which competitive contemporary literature.