Research Staff Member, IBM T. J. Watson Research Center
My research interest lies in data mining for biomedical domain with special interest in healthcare informatics and computational biology. I am broadly interested in developing novel data mining techniques to analyze the enormous amounts of data that is being generated in the form of individual clinical, genomic and genetic information. Effective integration of the partial information available in any of these genomic and clinical data can further help reveal disease complexities in greater detail. My interdisciplinary research focuses on analyzing the issues related to integrating diverse biomedical datasets and developing data mining approaches that try to address those challenges. During my PhD, I have worked on building integrative data mining approaches by combining multiple clinical and genetic datasets. In particular, my focus was primarily on (i) discovering relationships among the diverse clinical and genomic factors, (ii) enhancing the interpretability of integrative models by incorporating known medical knowledge, and (iii) developing more customized models to take the heterogeneity of samples into account. I developed novel data mining techniques based on multivariate predictive models, association analysis technique, and clustering. I also worked as research intern at Robert Bosch research and technology center for two summers. First time, I developed a clustering based feature extraction technique to extract useful feature from short-text based survey questions, which use both data similarity and the text similarity simultaneously to build new features. In the second term of my internship, I used association analysis technique based predictive model to account for the high numbers of missing values of the categorical data.