Discriminative, Generative and Imitative Learning
نویسندگان
چکیده
I propose a common framework that combines three different paradigms in machine learning: generative, discriminative and imitative learning. A generative probabilistic distribution is a principled way to model many machine learning and machine perception problems. Therein, one provides domain specific knowledge in terms of structure and parameter priors over the joint space of variables. Bayesian networks and Bayesian statistics provide a rich and flexible language for specifying this knowledge and subsequently refining it with data and observations. The final result is a distribution that is a good generator of novel exemplars. Conversely, discriminative algorithms adjust a possibly non-distributional model to data optimizing for a specific task, such as classification or prediction. This typically leads to superior performance yet compromises the flexibility of generative modeling. I present Maximum Entropy Discrimination (MED) as a framework to combine both discriminative estimation and generative probability densities. Calculations involve distributions over parameters, margins, and priors and are provably and uniquely solvable for the exponential family. Extensions include regression, feature selection, and transduction. SVMs are also naturally subsumed and can be augmented with, for example, feature selection, to obtain substantial improvements. To extend to mixtures of exponential families, I derive a discriminative variant of the ExpectationMaximization (EM) algorithm for latent discriminative learning (or latent MED). While EM and Jensen lower bound log-likelihood, a dual upper bound is made possible via a novel reverse-Jensen inequality. The variational upper bound on latent log-likelihood has the same form as EM bounds, is computable efficiently and is globally guaranteed. It permits powerful discriminative learning with the wide range of contemporary probabilistic mixture models (mixtures of Gaussians, mixtures of multinomials and hidden Markov models). We provide empirical results on standardized data sets that demonstrate the viability of the hybrid discriminative-generative approaches of MED and reverse-Jensen bounds over state of the art discriminative techniques or generative approaches. Subsequently, imitative learning is presented as another variation on generative modeling which also learns from exemplars from an observed data source. However, the distinction is that the generative model is an agent that is interacting in a much more complex surrounding external world. It is not efficient to model the aggregate space in a generative setting. I demonstrate that imitative learning (under appropriate conditions) can be adequately addressed as a discriminative prediction task which outperforms the usual generative approach. This discriminative-imitative learning approach is applied with a generative perceptual system to synthesize a real-time agent that learns to engage in social interactive behavior. Thesis Supervisor: Alex Pentland Title: Toshiba Professor of Media Arts and Sciences, MIT Media Lab Discriminative, Generative and Imitative Learning
منابع مشابه
Discriminative Models for Semi-Supervised Natural Language Learning
An interesting question surrounding semisupervised learning for NLP is: should we use discriminative models or generative models? Despite the fact that generative models have been frequently employed in a semi-supervised setting since the early days of the statistical revolution in NLP, we advocate the use of discriminative models. The ability of discriminative models to handle complex, high-di...
متن کاملSocratic Learning
Modern machine learning techniques often use discriminative models that require large amounts of labeled data. Since generating labeled training data sets is expensive, an alternative approach is to use a generative model, which leverages a simple heuristic to weakly label data. Domain experts prefer using generative models because they “tell a story” about their data. Unfortunately, generative...
متن کاملLearning Discriminative Metrics via Generative Models and Kernel Learning
Metrics specifying distances between data points can be learned in a discriminative manner or fromgenerative models. In this paper, we show how to unify generative and discriminative learning of met-rics via a kernel learning framework. Specifically, we learn local metrics optimized from parametricgenerative models. These are then used as base kernels to construct a global kerne...
متن کاملLearning Discriminative Sufficient Statistics Score Space for Classification
Generative score spaces provide a principled method to exploit generative information, e.g., data distribution and hidden variables, in discriminative classifiers. The underlying methodology is to derive measures or score functions from generative models. The derived score functions, spanning the so-called score space, provide features of a fixed dimension for discriminative classification. In ...
متن کاملDiscriminative Training of a Neural Network Statistical Parser
Discriminative methods have shown significant improvements over traditional generative methods in many machine learning applications, but there has been difficulty in extending them to natural language parsing. One problem is that much of the work on discriminative methods conflates changes to the learning method with changes to the parameterization of the problem. We show how a parser can be t...
متن کامل