Recapitulation of Ayurveda constitution types by machine learning of phenotypic traits

Recapitulation of Ayurveda constitution types by machine learning of phenotypic traits
— Read on journals.plos.org/plosone/article

You’ll need to have a basic understanding of the Ayurvedic concept of Prakriti to completely understand the relevance of this quote below. I’ll come back and add some commentary later. Suffice it to say that this type of study lends some credence to the reality of constitutional typing in traditional medical systems.

The present study was carried out on data of phenotypic attributes in 147 healthy individuals of three extreme Prakriti types, from a genetically homogeneous population of Western India. Unsupervised and supervised machine learning approaches were used to infer inherent structure of the data, and for feature selection and building classification models for Prakriti respectively. These models were validated in a North Indian population. Unsupervised clustering led to emergence of three natural clusters corresponding to three extreme Prakriti classes. The supervised modelling approaches could classify individuals, with distinct Prakriti types, in the training and validation sets. This study is the first to demonstrate that Prakriti types are distinct verifiable clusters within a multidimensional space of multiple interrelated phenotypic traits.