Big Data

In the area that deals with gargantuan amounts of data for the purpose of mining, machine learning and prediction, broadly known as “Big Data”, the Department of Computer Science is very heavily involved through the participation of a several faculty members in many interdisciplinary projects.

The work and expertise is in two categories: the first consists of general tools for handling, indexing and querying very large amounts of data, regardless of the nature of the data. The second comprises tools for special kinds of “big data”, such as images (optical, X-ray, fMRI, videos, text, genome data, etc.) and it is represented in the work of most faculty in the Computer Vision Laboratory and in the Center for Bioinformatics.

Such tools become indispensable with regard to the creation of ATLASes of various kinds, piecing together the structure of a particular object from millions of observations. Also, they become sophisticated procedures for mining information to create models for prediction, thus leading to diagnosis of events (such as disease, or the creation of a signal, and so on). Big Data Tools are among the latest tools in the service of the study of Brain and Behavior and are poised to revolutionize the field by discovering hidden relationships in the quantities involved in Brain and Behavior that up to now remained elusive.

We envision Big Data techniques being applied to: billions of frames in videos collected in rooms where children are in order to discover autistic behavior; billions of fMRI frames from a brain in action in order to discover dynamical system models underlying the activity of the neural circuits; billions of records from wearable sensors (including microscopic cameras) that will measure the calorie intake of an individual or other behaviors linked to neurodegenerative diseases. The list is endless and the only potential obstacle is the limit of our imagination.