Golay Code Transformations for Ensemble Clustering in Application to Medical Diagnostics

Clinical Big Data streams have accumulated largescale multidimensional data about patients’ medical conditions and drugs along with their known side effects. The volume and the complexity of this Big Data streams hinder the current computational procedures. Effective tools are required to cluster and systematically analyze this amorphous data to perform data mining methods including discovering knowledge, identifying underlying relationships and predicting patterns. This paper presents a novel computation model for clustering tremendous amount of Big Data streams. The presented approach is utilizing the error-correction Golay Code. This clustering methodology is unique. It outperforms all other conventional techniques because it has linear time complexity and does not impose predefined cluster labels that partition data. Extracting meaningful knowledge from these clusters is an essential task; therefore, a novel mechanism that facilitates the process of predicting patterns and likelihood diseases based on a semi-supervised technique is presented. Keywords—medical Big Data; clustering; machine learning; pattern recognition; prediction tool; Big Data classification; Golay