Influence functions and regularity tangents for efficient active learning

TL;DR

This paper introduces an efficient active learning method that uses influence functions and regularity tangents to identify the most informative data points for training regression models, enhancing data selection efficiency.

Contribution

The authors propose a novel technique combining influence functions and regularity tangents for rapid, effective active learning in regression tasks, with minimal additional storage.

Findings

The method computes a 'curiosity' measure for data points using a tangent vector and gradient inner product.

It is computationally efficient, adding only a constant slow-down during training.

The approach effectively guides data selection for improved model training.

Abstract

In this paper we describe an efficient method for providing a regression model with a sense of curiosity about its data. In the field of machine learning, our framework for representing curiosity is called Active Learning, which concerns the problem of automatically choosing data points for which to query labels in the semi-supervised setting. The methods we propose are based on computing a "regularity tangent" vector that can be calculated (with only a constant slow-down) together with the model's parameter vector during training. We then take the inner product of this tangent vector with the gradient vector of the model's loss at a given data point to obtain a measure of the influence of that point on the complexity of the model. In the simplest instantiation, there is only a single regularity tangent vector, of the same dimension as the parameter vector. Thus, in the proposed technique, once training is complete, evaluating our "curiosity" about a potential query data point can be done as quickly as calculating the model's loss gradient at that point. The new vector only doubles the amount of storage required by the model. We show that the quantity computed by our technique is an example of an "influence function", and that it measures the expected squared change in model complexity incurred by up-weighting a given data point. We propose a number of ways for using this and other related quantities to choose new training data points for a regression model.