UTML TR 2008 – 002 Learning and Evaluating Boltzmann Machines

@inproceedings{Salakhutdinov2008UTMLT2,
  title={UTML TR 2008 – 002 Learning and Evaluating Boltzmann Machines},
  author={Ruslan Salakhutdinov},
  year={2008}
}
We provide a brief overview of the variational framework for obtaining deterministic approximations or upper bounds for the log-partition fu ction. We also review some of the Monte Carlo based methods for estimating partiti on functions of arbitrary Markov Random Fields. We then develop an annealed impo rtance sampling (AIS) procedure for estimating partition functions of rest icted Boltzmann machines (RBM’s), semi-restricted Boltzmann machines (SRBM’s), an d Boltzmann machines (BM’s). Our… CONTINUE READING