Maintaining Performance with Less Data

TL;DR

This paper introduces a dynamic data reduction method for neural network training in image classification, significantly decreasing training time and environmental impact while maintaining accuracy.

Contribution

It presents a novel dynamic data reduction technique that reduces input data during training without sacrificing model performance.

Findings

Training time reduced by up to 50%

Carbon emissions decreased proportionally

Accuracy maintained despite data reduction

Abstract

We propose a novel method for training a neural network for image classification to reduce input data dynamically, in order to reduce the costs of training a neural network model. As Deep Learning tasks become more popular, their computational complexity increases, leading to more intricate algorithms and models which have longer runtimes and require more input data. The result is a greater cost on time, hardware, and environmental resources. By using data reduction techniques, we reduce the amount of work performed, and therefore the environmental impact of AI techniques, and with dynamic data reduction we show that accuracy may be maintained while reducing runtime by up to 50%, and reducing carbon emission proportionally.