Exploiting compositionality to explore a large space of model structures

10/16/2012
by   Roger Grosse, et al.
0

The recent proliferation of richly structured probabilistic models raises the question of how to automatically determine an appropriate model for a dataset. We investigate this question for a space of matrix decomposition models which can express a variety of widely used models from unsupervised learning. To enable model selection, we organize these models into a context-free grammar which generates a wide variety of structures through the compositional application of a few simple rules. We use our grammar to generically and efficiently infer latent components and estimate predictive likelihood for nearly 2500 structures using a small toolbox of reusable algorithms. Using a greedy search over our grammar, we automatically choose the decomposition structure from raw data by evaluating only a small fraction of all models. The proposed method typically finds the correct structure for synthetic data and backs off gracefully to simpler models under heavy noise. It learns sensible structures for datasets as diverse as image patches, motion capture, 20 Questions, and U.S. Senate votes, all using exactly the same code.

READ FULL TEXT
research
11/15/2020

Learning of Structurally Unambiguous Probabilistic Grammars

The problem of identifying a probabilistic context free grammar has two ...
research
12/20/2016

Grammar rules for the isiZulu complex verb

The isiZulu verb is known for its morphological complexity, which is a s...
research
08/07/2017

Generative Statistical Models with Self-Emergent Grammar of Chord Sequences

Generative statistical models of chord sequences play crucial roles in m...
research
01/27/2020

Unsupervised Program Synthesis for Images using Tree-Structured LSTM

Program synthesis has recently emerged as a promising approach to the im...
research
12/18/2022

Unsupervised Discontinuous Constituency Parsing with Mildly Context-Sensitive Grammars

We study grammar induction with mildly context-sensitive grammars for un...
research
08/31/2016

The Generalized Smallest Grammar Problem

The Smallest Grammar Problem -- the problem of finding the smallest cont...
research
06/03/2016

Scene Grammars, Factor Graphs, and Belief Propagation

We consider a class of probabilistic grammars for generating scenes with...

Please sign up or login with your details

Forgot password? Click here to reset