I am an applied statistician by training and am currently an associate professor in statistics. My research focuses on computational statistics with an emphasis on developing novel methodologies for analyzing datasets in challenging form. I have addressed problems in clustering of time series, hierarchical data, and text data. My work has been published in peer-reviewed statistics journals and led to national paper competition awards. On the application side, I have developed a forecasting algorithm jointly with colleagues in electrical engineering. With unstructured data, I worked on building clustering, classification, and sentiment analysis models. In the health sciences, I have worked on mixture modeling, spatial clustering, and forecasting projects on datasets from Electronic Health Records and publicly available health data.
Office Phone: +1-605-688-6316
Email: semhar[dot]michael[at]sdstate[dot]edu
Associate Professor of Statistics
Associate Editor for the Journal of Classification
Education
Ph.D. University of Alabama (2015), Applied Statistics
MS. University of North Dakota (2011), Mathematics
BS. University of Asmara, Eritrea (2006), Applied Mathematics
Research Interests
Computational statistics, Mixture modeling, unsupervised learning, clustering, unstructured data analysis, forensic source identification problems