Knowledge-driven Active Learning

by   Gabriele Ciravegna, et al.

In the last few years, Deep Learning models have become increasingly popular. However, their deployment is still precluded in those contexts where the amount of supervised data is limited and manual labelling expensive. Active learning strategies aim at solving this problem by requiring supervision only on few unlabelled samples, which improve the most model performances after adding them to the training set. Most strategies are based on uncertain sample selection, and even often restricted to samples lying close to the decision boundary. Here we propose a very different approach, taking into consideration domain knowledge. Indeed, in the case of multi-label classification, the relationships among classes offer a way to spot incoherent predictions, i.e., predictions where the model may most likely need supervision. We have developed a framework where first-order-logic knowledge is converted into constraints and their violation is checked as a natural guide for sample selection. We empirically demonstrate that knowledge-driven strategy outperforms standard strategies, particularly on those datasets where domain knowledge is complete. Furthermore, we show how the proposed approach enables discovering data distributions lying far from training data. Finally, the proposed knowledge-driven strategy can be also easily used in object-detection problems where standard uncertainty-based techniques are difficult to apply.


page 2

page 4

page 7

page 13

page 14


Can Domain Knowledge Alleviate Adversarial Attacks in Multi-Label Classifiers?

Adversarial attacks on machine learning-based classifiers, along with de...

Adversarial Sampling for Active Learning

This paper describes ASAL a new active learning strategy that uses uncer...

How useful is Active Learning for Image-based Plant Phenotyping?

Deep learning models have been successfully deployed for a diverse array...

Generation of Near-Optimal Solutions Using ILP-Guided Sampling

Our interest in this paper is in optimisation problems that are intracta...

Take a Chance: Managing the Exploitation-Exploration Dilemma in Customs Fraud Detection via Online Active Learning

Continual labeling of training examples is a costly task in supervised l...

Minimizing Supervision in Multi-label Categorization

Multiple categories of objects are present in most images. Treating this...

DADO – Low-Cost Selection Strategies for Deep Active Design Optimization

In this experience report, we apply deep active learning to the field of...

Please sign up or login with your details

Forgot password? Click here to reset