Research Areas
List View
Evolutionary Computation
Evolutionary Computation is a biologically inspired machine learning method that aims to solve (or optimize) complex problems by performing an intelligent parallel search in the solution space. Our re...

        •Neuroevolution
Neuroevolution is a method for optimizing neural network weights and topologies using evolutionary computation. It is particularly useful in sequential decision tasks that are partially observable (i....

        •Theory of Evolutionary Computation
Our work focuses on applying measure theory and martingale analysis to develop new evolutionary algorithms with known properties, as well as a theoretical characterization, performance measures, and c...

        •Multiobjective Optimization
Instead of finding a single optimal solution to any given problem, multiobjective methods aim at finding a Pareto-front, which represents all of the trade-offs between objectives within the domain. A ...

Cognitive Science
Cognitive Science attempts to build computational models of human (and animal) cognitive processes in order to improve our understanding of natural intelligent systems. Such an understanding then form...

        •Natural Language Processing (Cognitive)
Our research in natural language processing focuses on cognitive modeling. In particular, it aims at bridging the gap between subsymbolic representations and complex high-level behavior. The models ar...

        •Memory
Episodic memory is the record of events that the individual experiences throughout his/her/its life. The events are stored temporarily in the hippocampus and related areas immediately as they happen,...

        •Concept and Schema Learning
An agent can truly understand the meaning of its knowledge structures, and utilize them most effectively, only if that knowledge is grounded on sensorimotor interactions with the world. We aim at buil...

        •Brain and Cognitive Disorders
The goal is to develop computational insight into how damage to the brain occurs and how it can be mitigated and repaired. Our work focuses in particular in disorders such as schizophrenia, aphasia, a...

Computational Neuroscience
A computational model is a complete description of how a neural system functions, and in that sense the ultimate specification of neuroscience theory. The models are constrained by and validated with...

        •Visual Cortex
Our goal is to understand the development and function of the visual cortex through computational modeling. The main idea is that the visual cortex is a continuously-adapting structure in a dynamic eq...

        •Memory
Episodic memory is the record of events that the individual experiences throughout his/her/its life. The events are stored temporarily in the hippocampus and related areas immediately as they happen,...

        •Neuroimaging
Neuroimaging techniques such as fMRI, EEG, and VSDI allow visualizing brain structure and function. Computational techniques are developed to analyze the results, and in some cases to model the underl...

        •Brain and Cognitive Disorders
The goal is to develop computational insight into how damage to the brain occurs and how it can be mitigated and repaired. Our work focuses in particular in disorders such as schizophrenia, aphasia, a...

Other Areas
This category includes work that contributes to areas other than those specifically identified in our area hierarchy, including the general subareas listed below, as well as competitive multiagent sea...

        •Supervised Learning
In supervised learning the desired outputs are known for each input, and the task is to learn a mapping between them that generalizes well to new inputs. Our research in this area includes various app...

        •Reinforcement Learning
In reinforcement earning tasks the desired behavior is not known; only sparse feedback on how well the agent is doing is provided. Reinforcement learning techniques include value-function and policy i...

        •Unsupervised Learning, Clustering, and Self-Organization
Unsupervised learning does not require annotation or labeling from a human teacher; the idea is to learn the structure of the data from unlabeled examples. The most common unsupervised learning task ...

        •Algorithm Portfolios
Algorithm portfolio methods operate in problem domains for which there are multiple algorithms with complementary strengths. The portfolio method applies patterns learned from experience to better all...

Applications
Much of our work on applications involves neuroevolution of behavior in real-world domains such as control, robotics, game playing, and artificial life, but also design and optimization of wavelets, s...

        •Control
Control theory and engineering study how to produce the desired behavior in a variety of dynamical systems, e.g. the electro-mechanical systems in robots and the chemical systems in manufacturing plan...

        •Robotics
Our work in robotics focuses on neuroevolution of intelligent locomotion and navigation, as well as cooperation among multiple robots. We have also worked on learning sensorimotor embeddings through u...

        •Artificial Life
Artificial life is a broad field that involves imitating life in software, hardware, and biochemistry. Our work in this area focuses on computational modeling of behavior in simulated agents and teams...

        •Game Playing
Games constitute an effective platform for developing and testing artificial intelligence techniques: they are well defined and easy to implement, yet challenging and fun. Our work in this area involv...

        •Satisfiability
The problem of propositional satisfiability (SAT) is the classic NP-complete problem. It asks whether a Boolean expression is satisfiable: whether an assignment of Boolean values to its variables exis...