“Fortune at the Bottom of the Classifier Pyramid”: A Novel Approach to Human Activity Recognition

“Fortune at the Bottom of the Pyramid”, penned by the noted economist Prof. C. K. Prahalad1, talked about the wealth creating potentials of entities at the bottom of the economic hierarchy. Taking the same cue into classifier research in Machine Learning, Naive Bayes classifier is often the beaten boy among its contemporaries. In this work, Naive Bayes (NB), is revisited by looking at heuristic ways of improving any skewed data bias, systematic and weighted magnitude errors; to address the problem of human activity recognition and propose the Improved Naive Bayesian Algorithm. The novelty of this work is in adjusting the human activity recognition problem as a special case of text classification.