Rami Mahdi ~ Postdoctoral Associates



Office:
1305 York Avenue, 13th Floor
New York, NY 10021

Contact:
Phone: 502-759-9293
Email: ramimahdi at yahoo dot com

Research Interests:
Research and development of interpretable models such as clustering analysis, penalized learning, and minimal hyper basis function networks (Reduced HyperBF). These tools possess the potential to provide evidence for causation analysis and to uncover subpopulations information. Such information is invaluable in the areas of bioinformatics and artificial intelligence. For example, in bioinformatics, it can provide evidence for biomarkers and uncover information about disease’ subtypes. In artificial intelligence, such models are needed for knowledge extractions.


 
Papers:
  • Mahdi R., Mezey J. 2010. Sub-local constraint based learning of Bayesian networks with a mutual dependence criterion. submitted.
  • Rami N. Mahdi and Eric C. Rouchka, “Reduced HyperBF Networks: Regularization by Explicit Complexity Reduction and Scaled Rprop Based Training”, Under revision for IEEE Transactions on Neural Networks. Link
  • Rami N. Mahdi and Eric C. Rouchka, “Feature selection in cancer classification from mRNA data based on localized dimension reduction”, Proceedings of the Seventh International Conference on Machine Learning and Applications (ICMLA 2009), pp. 443-448, December 12-15, 2009, Miami, Florida.
  • Rami N. Mahdi and Eric C. Rouchka, “Model based unsupervised learning guided by abundant background samples”, Proceedings of the Seventh International Conference on Machine Learning and Applications (ICMLA 2008), pp. 203-210, December 11-13, 2008, San Diego, California. Link
  • Rami N. Mahdi and Eric C. Rouchka, “RBF-TSS: Identification of Transcription Start Site in Human Using Radial Basis Functions Network and Oligonucleotide Positional Frequencies”, PloS One, Vol. 4, No. 3., e4878, 2009. Link
  • Rami N. Mahdi and Eric C. Rouchka, “Evidence of bias towards buffered codons in human transcripts “, Proceedings of the Eighth IEEE Symposium on Signal Processing and Information Technology (ISSPIT 2008), pp. 29-34, December 16-19, 2008, Sarajevo, Bosnia and Herzegovina. Link
  • Hichem Frigui, Rami N. Mahdi, Meredith, J., “Combining feedback and image database categorization in CBIR”, Proceedings of the 2008 IEEE International Conference on Multimedia and Expo, pp. 1277-1280, June 23-28, 2008, Hanover, Germany.
  • Hichem Frigui, Rami N. Mahdi, “Semi-supervised clustering and feature discrimination with instance-level constraints”, In Proceedings of the IEEE International Fuzzy Systems Conference, pp. 1-6, July 23-26, 2007.