Collaborative Support Faculty

Thayasivam Umashanger

Associate Professor, Mathematics
thayasivam@rowan.edu
Website

Education

BSc (Statistics), University of Colombo
MS (Statistics), University of Georgia
PhD (Statistics), University of Georgia

Research Expertise

Statistical Data Mining and Bio marker discovery)| Robust Estimation| statistical application in stem

My interdisciplinary statistical research spanned diverse areas including statistical data mining, speaker recognition, peak detection and network security. My current projects include analyzing pharmaceutical data with data mining, speaker identification, and spoof detection and network security.

I also am performing Biomarker research aimed at optimizing and verifying the utility of autoantibody biomarkers for early diagnosis-Biomarker Discovery Center at Rowan SOM. Where I ensure that all of the data evaluation strategies and methodologies employed in the studies will take full advantage of any recent developments, improvements and alternative analytical approaches.

Memberships

American Statistical Association
Institute of Mathematical Statistics
Institute of Applied Statistics Sri Lanka

Recent Publications

  1. Isaac B. Muck, Vasil Hnatyshin, and Umashanger Thayasivam, Accuracy of Class Prediction using Similarity Functions in PAM, in Proc. of IEEE International Conference on Industrial Technology (IEEE ICIT 2016), Taipei, Taiwan, March 14-17, 2016
  2. Umashanger, T.; Kuruwita, C.; Ramachandran, R.P., “Robust L2E Parameter Estimation of Gaussian Mixture Models: Comparison with Expectation Maximization,” Neural Information Processing,LNCS9489, DOI: 10.1007/978-3-319-26555-1-32, November 2015.
  3. William Ezekiel.; Umashanger, T., “A Comparison of Supervised Learning Techniques for 
Clustering,” Neural Information Processing, ISBN 978-3-319-26555-1, November 2015.
  4. DeMarshall CA, Han M, Ngele EP, Sarkar A, Archarya NK. Godsey G, Goldwaser El, Kosciuk M; Thayasivam U, Belinka B, Nagele RG and the Parkinson’s Study Group Investigators. “Potential utility of autoantibodies as blood-based biomarkers for early detection and diagnosis of Parkinson ‘s disease”. Immunol Lett. 168(1) 80-88. PMID:26386375,2015.
  5. S. Hnatyshyn, U. Thayasivam, V. Hnatyshin and C. White. Chapter: 7, Machine learning algorithms for metabolomics applications. Identification and Data Processing Methods in Metabolomics, Chapter: 7. Future Science Book Series, pp. 96-110 DOI: 10.4155/fseb2013.14.163, 2015.

One Medical Center Drive
Stratford, New Jersey 08084-1501
Technical contact for the web site
Privacy Statement