Optimization and Learning Center

Research

The Center draws on strong faculty expertise in developing, and implementing into software, core optimization and statistical learning technologies in the form of novel mathematical models and algorithms.

Faculty research is focused on the following interconnected areas:

Optimization

In the context of statistical and machine learning, optimization discovers the best model for making predictions given the available data. This powerful paradigm has led to major advances in speech and image recognition—and the number of future applications is expected to grow rapidly. More generally, numerical optimization finds the best design or operating condition of a complex system. Researchers collaborate with a wide range of data scientists to design new algorithms that will play a central role in the creation of more intelligent systems.

Statistics and machine learning

With the advent of improved technology for storage and collection of data, we now have access to more information than ever before. Statistics and Machine Learning involve the process of using the accumulated data to make inferences and provide insights. Models and algorithms make it possible to conduct explorative analysis, expose trends and make predictions from the collected data. Recent applications range from Web search engines to media marketing.

Applications

As better statistical models are informed by larger datasets and are coupled with better optimization algorithms, it becomes possible to obtain massively improved outcomes on many difficult problems. This includes problems in healthcare systems, natural language processing, video & audio processing, and critical infrastructure including electric power grids & communication networks. Each problem domain has its own challenges that need to be addressed including data collection & pre-processing, model selection & post-processing, and optimal design & control of the system.