2015 NEXTASystemforRealWorldDevelopm

• (Jamieson et al., 2015) ⇒ Kevin Jamieson, Lalit Jain, Chris Fernandez, Nick Glattard, and Robert Nowak. (2015). “NEXT: A System for Real-world Development, Evaluation, and Application of Active Learning.” In: Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 2.

Subject Headings: Active Learning System, NEXT.

Cited By

• http://scholar.google.com/scholar?q=%222015%22+NEXT%3A+A+System+for+Real-world+Development%2C+Evaluation%2C+and+Application+of+Active+Learning
• http://dl.acm.org/citation.cfm?id=2969442.2969536&preflayout=flat#citedby

Quotes

Abstract

Active learning methods automatically adapt data collection by selecting the most informative samples in order to accelerate machine learning. Because of this, real-world testing and comparing active learning algorithms requires collecting new datasets (adaptively), rather than simply applying algorithms to benchmark datasets, as is the norm in (passive) machine learning research. To facilitate the development, testing and deployment of active learning for real applications, we have built an open-source software system for large-scale active learning research and experimentation. The system, called NEXT, provides a unique platform for real-world, reproducible active learning research. This paper details the challenges of building the system and demonstrates its capabilities with several experiments. The results show how experimentation can help expose strengths and weaknesses of active learning algorithms, in sometimes unexpected and enlightening ways.

1 Introduction

We use the term “active learning” to refer to algorithms that employ adaptive data collection in order to accelerate machine learning. By adaptive data collection we mean processes that automatically adjust, based on previously collected data, to collect the most useful data as quickly as possible. This broad notion of active learning includes multi-armed bandits, adaptive data collection in unsupervised learning (e.g. clustering, embedding, etc.), classification, regression, and sequential experimental design. Perhaps the most familiar example of active learning arises in the context of classification. There active learning algorithms select examples for labeling in a sequential, data-adaptive fashion, as opposed to passive learning algorithms based on preselected training data.

---

---

References

;

	Author	volume	Date Value	title	type	journal	titleUrl	doi	note	year
2015 NEXTASystemforRealWorldDevelopm	Kevin Jamieson Lalit Jain Chris Fernandez Nick Glattard Robert Nowak			NEXT: A System for Real-world Development, Evaluation, and Application of Active Learning						2015