People
Faculty
Artur Dubrawski
Director (Auton Lab), Systems Scientist (Robotics Institute), and Adjunct Professor (Heinz College)
Biography
Artur Dubrawski considers himself a scientist and a practitioner. He has been tainted with real world entrepreneurial experiences. He had started up a successful company specializing in integration and deployment of advanced control systems and technological devices. He had also been affiliated with startups incorporated by others: Schenley Park Research, a data mining consultancy and a CMU spin-off, where he was a scientist; and Aethon, a company building robots to automate transportation in hospitals, where he served as a Chief Technical Officer. Artur returned to CMU in 2003 to rejoin the Robotics Institute's Auton Lab. He works on a range of applied computer intelligence endeavors, and he teaches data mining and business intelligence to graduate students at the CMU Heinz College School of Information Systems and Management. In his previous academic life, he pursued machine learning approaches to mobile robot navigation and control, and other applications of adaptive autonomous systems. Artur received a Ph.D. in robotics and automation from the Institute of Fundamental Technological Research, Polish Academy of Sciences, and a M.Sc. in aircraft engineering from Warsaw University of Technology. In 1995/96 he spent a year at CMU as a visiting Fulbright scholar. In January 2006 Artur Dubrawski was named the director of the Auton Lab.
Research Interests
I am interested in intelligent systems that work, are useful, and make economic sense, and in finding ways to effectively build and deploy them. My work is driven by real-world applications, currently in the areas of public health, food safety, nuclear safety and health of equipment. It involves researching new machine learning algorithms and data structures to facilitate probabilistic modeling, predictive analysis, interactive exploration, and understanding of data.
Tags
Active Learning, Applications, Association Rules, Biosurveillance, Dynamic Social Networks, Food Safety, GDA, Health of Equipment, Link Analysis, Locally Weighted Learning, Memory-based Learning, Mixture Models, Nuclear Safety, Optimization, Social Networks
Recent Papers
-
Learning Compressible Models
(2010)
Learning Compressible Models -
Trade-offs between Agility and Reliability of Predictions in Dynamic Social Networks Used to Model Risk of Microbial Contamination of Food
(2009)
Best paper award, ASONAM 2009 -
T-Cube Web Interface in Support of Real-Time Bio-surveillance Program
(2009)
-
Evolution of a Useful Autonomous System
(2009)
-
Learning the Semantic Correlation: An Alternative Way to Gain from Unlabeled Text
(2008)
A semi-supervised learning algorithm on text - (7 more)
Software
-
AFDL (Activity From Demographics and Links)
Predicting activity of entities from linkages between entities and their demographics -
Vizier
Old but fast locally weighted regression for Windows. -
TCWI (T-Cube Web Interface)
Front end for interactive visualization, navigation, and analysis of multidimensional time series data
Jeff Schneider
Associate Research Professor
Biography
Dr. Jeff Schneider is an associate research professor in the Carnegie
Mellon University School of Computer Science. He received his PhD in
Computer Science from the University of Rochester in 1995. He has over 15
years experience developing, publishing, and applying machine learning
algorithms in government, science, and industry. He has dozens of
publications and has given numerous invited talks and tutorials on the
subject.
Dr. Schneider was the co-founder and CEO of Schenley Park Research,
Inc. (SPR), a company dedicated to bringing new machine learning algorithms
to industry. Later, he developed a new machine-learning based CNS drug
discovery system and spent a two-year sabbatical as the Chief Informatics
Officer of Psychogenics, Inc. to set up and commercialize the system. Through
his work at CMU and his commercial and consulting efforts, he has worked
with several dozen companies and government agencies including ten Fortune
500 companies, and groups from around the world.
Research Interests
My research is in learning, optimization, data mining, scheduling, robotics, and intelligent control. I am also interested in related areas including navigation, manipulation, vision, and other types of sensing. My efforts are focused on the application of these methods to real-world scientific, commercial, and government problems.
Tags
Active Learning, Applications, Astrostatistics, Bayesian Networks, Biosurveillance, Cached Sufficient Statistics, Efficient Statistical Algorithms, GDA, Kd-trees and Ball-trees, Link Analysis, Locally Weighted Learning, Markov Decision Processes, Memory-based Learning, Optimization, Spatial Statistics, Statistical Data Mining for Astrophysics, WSARE
Recent Papers
-
Copula-based Kernel Dependency Measures
(2012)
Copula-based Kernel Dependency Measures -
Nonparametric Kernel Estimators for Image Classification
(2012)
nonparametric kernel estimation for image classification -
Support Distribution Machines
(2012)
Kernel algorithms on distributions -
Nonparametric Estimation of Conditional Information and Divergences
(2012)
nonparametric conditional mutual information estimator -
Learning Auto-regressive Models from Sequence and Non-sequence Data
(2011)
Combining sequence and non-sequence data to improve dynamic model learning - (52 more)
Software
-
cGraph
CGraph is an algorithm to quickly learn a graph-based model of the underlying connections of a set of entities given link data. -
k-groups/GDA
The group detection algorithm (GDA) finds underlying groupings of entities given a set of observed links and demographic information. -
Vizier
Old but fast locally weighted regression for Windows. -
Group Anomaly Detection Software
Matlab code of probabilistic models for group anomaly detection. -
Nonparametric Divergence Estimation
C++ and pure Matlab implementations of nonparametric divergence estimation. -
Source Code: Multi-Label Output Codes using Canonical Correlation Analysis
Multi-label Output Code using Canonical Correlation Analysis -
Active Model Fitting
Active Model Fitting