Personalization of Deep Learning

TL;DR

This paper explores techniques for personalizing deep learning models by using curriculum learning and data augmentation, improving individual performance but potentially reducing general dataset accuracy.

Contribution

It introduces specific curriculum learning strategies and data grouping methods for personalization, demonstrating their effectiveness in enhancing individual model performance.

Findings

Personalization improves individual data performance

Data augmentation can reduce general dataset accuracy

Curriculum learning strategies are effective for personalization

Abstract

We discuss training techniques, objectives and metrics toward personalization of deep learning models. In machine learning, personalization addresses the goal of a trained model to target a particular individual by optimizing one or more performance metrics, while conforming to certain constraints. To personalize, we investigate three methods of ``curriculum learning`` and two approaches for data grouping, i.e., augmenting the data of an individual by adding similar data identified with an auto-encoder. We show that both ``curriculuum learning'' and ``personalized'' data augmentation lead to improved performance on data of an individual. Mostly, this comes at the cost of reduced performance on a more general, broader dataset.