Milos Hauskrecht

Associate professor of Computer Science

Address:

Department of Computer Science

5329 Sennott Square

University of Pittsburgh

Pittsburgh, PA 15260

412-624-8845 (voice)

412-624-8854 (fax)

e-mail: milos_at_cs.pitt.edu

Current positions/openings:

A postdoctoral research associate position in Machine Learning and applications of ML in medicine

Education:

PhD in Computer Science, Department of Electrical Engineering and Computer Science, Massachusetts Institute of Technology, Cambridge, MA, August 1997.

Research:

My research work is in machine learning, data mining, probabilistic modeling and their applications in biomedical informatics. The main focus of my current research is the development of modern machine learning technologies for analysis of high dimensional time-series data in electronic health record (EHR) data and their application in clinical decision support and clinical alerting.

Publications

Current research projects:

Outlier detection in large clinical databases. Development of statistical outlier detection methods for identification of unusual outcomes and patient management decisions. Application to alerting in ICU data streams.
- Homer Warner Award for the work on the outlier-based clinical alerting at AMIA 2010
  M. Hauskrecht, M. Valko, I.Batal, G. Clermont, S. Visweswaran, G. Cooper. Conditional Outlier Detection for Clinical Alerting, Annual American Medical Informatics Association (AMIA) Symposium , November 2010.
Mining Temporal Predictive Patterns.
Active learning for intelligent sample collection .
Markov decision processes. Design of efficient value-function and policy approximation methods for solving large MDPs. Extensions of MDPs to continuous time, continuous state and partially observable settings.

Current research funding:

NIH/NIGMS. R01. Detecting deviations in clinical care in ICU data streams (Principal Investigator)
NIH/NLM. R01. Using medical records repositories to improve the alert system design. (Principal Investigator)
NSF. Discovering Complex Anomalous Patterns (co-investigator, PI: Dubrawski, CMU and Cooper, Uiniversity of Pittsburgh)

Other recent projects:

Modeling, analysis and optimization of stochastic networks Noisy-or component analysis. Modeling and learning of transportation networks. Resource optimizations for transportation networks.
Machine learning methods for proteomics and genomics. Statistical methods for identification of hidden regulatory pathways from gene expression data. Algorithms for preprocessing and analysis of high-throughput MS proteomic and genomic data and for biomarker discovery. Methods for protein ID in whole sample proteomics using prior knowledge

Teaching:

CS 2750 Machine Learning

Current students

Saeed Amizadeh
Iyad Batal
Eric Heim
CharmGil Hong
Zitao Liu
Quang Nguyen

Shuguang Wang

Postdocs

Hamed Valizadegan
Lei Wu

Former students:

Branislav Kveton, PhD, Fall 2006, research scientist, Technicolor
Tomas Singliar, PhD, Fall 2008, research scientist, Boeing Research
Richard Pelikan, PhD Spring 2011, postdoc, USC
Michal Valko, PhD Summer 2011, postdoc, INRIA France
MS students:

David Krebs, MSc, Fall 2011
Eric Heim, MSc, Fall 2010
Adi Nemlekar, MSc, Summer 2007
Jose Nunez-Varela, MSc, Spring 2006
Aaron Cois, MSc, Spring 2006
Xinghua Lu, MSc, 2004
Elizabeth Clause, MSc
Gregory Nilsen, MSc
Past courses taught:

CS 1571 Introduction to Artificial Intelligence (Fall 2010, Fall 2007, Fall 2006, Fall 2003, Fall 2002, Fall 2001)
CS 441 Discrete mathematics for Computer Science ( Spring 2006, Spring 2005, Fall 2004)

CS 2710 Foundations of Artificial Intelligence (Fall 2005, Fall 2004)

CS 3710 Probabilistic Graphical Models (Advanced Topics in AI) (Fall 2005)

CS 2750 Machine Learning ( Spring 2010 , Spring 2007 , Spring 2004 , Spring 2003)

CS 3750 Advanced Machine Learning (Fall 2007)

CS 2740 Knowledge Representation (Fall 2008, Spring 2007)
CS 2001 Research topics in Computer Science. Module: Bayesian belief networks (Fall 2002, Fall 2001)

Last updated by milos on 05/30/07.