Spectral Algorithms for Learning Latent Variable Models

We are happy to share with you a recent talk by Sham Kakade from Microsoft recorded at the NYC Machine Learning meetup . In this talk he discusses a general and (computationally and statistically) efficient parameter estimation method for a wide class of latent variable models---including Gaussian mixture models, hidden Markov models and latent Dirichlet allocation---by exploiting a certain tensor structure in their low-order observable moments.

 

Sham Kakade is an Engineer at Microsoft

Come as you are. Do what you love. At Microsoft we help people and businesses throughout the world realize their full potential. We make this simple mission come to life every day through our passion to create technologies and develop products that touch just about every kind of customer. Working at Microsoft gives you the opportunity to do things that make a real difference in millions, even billions, of lives. To reach your full potential. So why not take a closer look at Microsoft? We think you’ll find that amazing things really do happen here.