frustratingly easy domain adaptation

6
Frustratingly Easy Domain Adaptation Hal Daume III

Upload: hu-ferrell

Post on 30-Dec-2015

26 views

Category:

Documents


1 download

DESCRIPTION

Frustratingly Easy Domain Adaptation. Hal Daume III. Introduction. Task: Developing Learning Algorithms that can be easily ported from one domain to another. Example: from newswire to biomedical docs. particularly interesting in NLP. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Frustratingly Easy Domain Adaptation

Frustratingly Easy Domain Adaptation

Hal Daume III

Page 2: Frustratingly Easy Domain Adaptation

Introduction

• Task: Developing Learning Algorithms that can be easily ported from one domain to another. Example: from newswire to biomedical docs.

• particularly interesting in NLP.• Idea: Transforming the domain adaptation learning

problem into a standard supervised learning problem to which any standard algorithm may be applied (eg., maxent, SVM)

• Transformation is simple – Augment the feature space of both the source and target data and use the result as input to a standard learning algorithm.

Page 3: Frustratingly Easy Domain Adaptation

Problem Formalization

Notation:• X the input space (typically either a real vector or a

binary vector) and Y the output space.• Ds to denote the distribution over source examples and

Dt to denote the distribution over target examples.• we have access to a samples Ds D∼ s of source

examples from the source domain, and samples Dt D∼ t of target examples from the target domain.

• assume that Ds is a collection of N examples and Dt is a collection of M examples (where, typically, N M).≫

• Goal: to learn a function h : X → Y with low expected loss with respect to the target domain.

Page 4: Frustratingly Easy Domain Adaptation

Adaptation by Feature Augmentation

• Take each feature in the original problem and make three versions of it: a general version, a source-specific version and a target-specific version.

• Augmented source data = General and source specific• Augmented Target data = General and target specific

Page 5: Frustratingly Easy Domain Adaptation

Results

• Tasks (see paper)

Page 6: Frustratingly Easy Domain Adaptation

Experimental Results

• See paper