Machine Learning

(Andrew Barto, Rod Grupen, David Jensen, David Kulp, Erik Learned-Miller, Sridhar Mahadevan, Andrew McCallum, Robbie Moll, Hava Siegelmann, Paul Utgoff)

Machine learning is the computational study of pattern discovery and skill acquisition. This includes methods by which artificial agents can improve their behavior while interacting with their environments, for example, by learning effective behavioral strategies from experience or by improving the knowledge structures forming the basis of their decisions. Machine learning also includes data mining techniques for finding patterns in large bodies of data. Specific research topics in the department include learning conceptual structures through developmental processes; improving control of stochastic and nonlinear dynamic systems through reinforcement feedback; learning robot control strategies; finding patterns in large bodies of data represented in graphical form, including social networks; extracting or retrieving information in natural language; classification of genetic data; and using learning methods for improving discrete optimization algorithms. Much of the machine learning research in the department is multi-disciplinary, with strong ties to research in statistics, operations research, cognitive and developmental psychology, neuroscience, and philosophy.

Autonomous Learning Laboratory
The Autonomous Learning Laboratory (ALL), formerly the Adaptive NetWorks (ANW) Laboratory, focuses on both machine and biological learning. Areas of study include reinforcement learning, artificial neural networks, and biologically-inspired models of adaptive motor control.

Biologically Inspired Neural & Dynamical Systems Laboratory
The Biologically Inspired Neural & Dynamical Systems Laboratory aims to apply techniques developed in computer science to problems in biology and to turn insights gained from biological systems to construct better computational algorithms. A specific goal is to employ computational techniques from machine learning, such as clustering and Bayesian network modeling, to solve problems in functional genomics. Another goal of the lab is to build mathematical models of neural circuitries in the brain.

Computational Biology Laboratory
The Computational Biology Laboratory uses a wide range of computational approaches to investigate fundamental biological problems in genetics, genomics, and structural molecular biology. These include regulatory network inference, the genetics of alternative splicing, protein structure prediction, and protein docking. We also work on technical improvements in variation detection and sequencing using DNA microarrays.

Information Extraction and Synthesis Laboratory
The Information Extraction and Synthesis Laboratory (IESL) specializes in the theoretical development and implementation of systems for extracting databases from unstructured text on the Web, and mining those databases to find patterns, predict the future, and provide decision support.

Knowledge Discovery Laboratory
KDL investigates how to find useful patterns in large and complex databases. We study the underlying principles of data mining algorithms, develop innovative techniques for knowledge discovery, and apply those techniques to practical tasks in areas such as fraud detection, scientific data analysis, and web mining.

Laboratory for Perceptual Robotics
The Laboratory for Perceptual Robotics investigates planning and control methodologies for complex, multi-objective robotic systems, geometric reasoning for automated assembly planning, and robot learning. Research platforms include integrated hand/arm systems, mobile robots, legged systems, and articulated stereo heads.

Machine Learning Laboratory
The Machine Learning Laboratory studies computational methods that enable machines to learn from experience. Our focus is on the problem of learning the representations on which more widely studied learning algorithms depend. Currently, we are investigating methods for online simultaneous acquisition and organization of deep networks of building block predicates and functions. Older work includes the ITI incremental decision tree induction programs.

Theoretical Computer Science Group
Theoretical Computer Science is the quantitative and formal study of computing: which problems can be solved? what resources (for example, time or memory space) are required to solve them? Our faculty specializing in a variety of areas, including the complexity of algebraic computations, the complexity of parallel computation, the descriptive complexity of computation, and the theory of parallel and distributed processing.