0:00:07my name is tone each of are i mean assistant professor computer science i been
0:00:11at columbia for a little over five years now and might area of research is
0:00:15machine learning i direct the company machine learning lab and you have a large group
0:00:20of students doing several really exciting projects the machine learning lab is really about this
0:00:26velocity of combining complication computer science
0:00:30and marrying into statistics because there is so much data out there where we have
0:00:35not just an information age of an information overload age and the real hopes to
0:00:41use computers to help us make sense of the data automatically
0:00:45we want them to learn much with people learn and also work with the types
0:00:49of data that we care about so the text we read the images we see
0:00:53we want computers to be able understand that you to that we're generating everyday at
0:00:58a faster and faster rate
0:01:00in biology for example there's millions of variables extends of thousand genes and it's almost
0:01:06impossible to have someone look at this data and come up with a theory about
0:01:10how the biology works
0:01:12so increasingly side to succumbing to computer scientists and machine order saying
0:01:17we've made all these measurements there's just too many variables and machine learning is one
0:01:21of the few tools that really can work with this type of data machine learning
0:01:25can provide us with a network description of visualisation clustering production and so on which
0:01:32sciences finding very valuable these days
0:01:35another thing we've been working with its social network analysis and
0:01:39biological networks is another natural counterpart
0:01:42looking at networks of proteins figure out how they interact with the proteins functions are
0:01:47how expression levels vary over time i think machine learning is one of these
0:01:52unusually lucky feels in that
0:01:55the foundations it's working from are useful to many other disciplines
0:02:01are particular research is machine learning applied to really complicated problems and datasets
0:02:08where there is some additional structure that space so images transform in various ways if
0:02:14you see picture someone it rotates the latter the right or you move around somebody's
0:02:18face in an image you still recognizing and so we're trying to incorporate that same
0:02:23type of
0:02:25structure into all are machine learning algorithms we also design algorithms and machine learning models
0:02:31at work on sequences so then we can handle things like a string of text
0:02:35we've been able to do that very successful using machine learning by modeling the sequence
0:02:40structure of the text
0:02:43give me two documents are not tell you they were uttered by the same person
0:02:47or there are some subtle stylistic things between those two documents that say that this
0:02:51is in the same person
0:02:53in my mind that's the really exciting future direction from machine learning one of the
0:02:57areas we concentrate on my group which is how to incorporate some of this invariance
0:03:02we know exist special you picture a short you tell to you still recognise that
0:03:07actually see later on
0:03:09and that's kind of the key i think too many real-world problems that
0:03:15invariance for that particular problem