Stop overkilling simple tasks with black-box models and use transparent models instead

TL;DR

This paper advocates replacing complex black-box deep learning models with transparent models for simple tasks to improve interpretability without sacrificing performance.

Contribution

It introduces a framework or methodology for substituting deep learning models with transparent alternatives in simple tasks, emphasizing interpretability.

Findings

Transparent models perform comparably to deep learning on simple tasks.

Using transparent models enhances interpretability and reduces complexity.

Deep learning excels in feature extraction but may be unnecessary for simple tasks.

Abstract

In recent years, the employment of deep learning methods has led to several significant breakthroughs in artificial intelligence. Different from traditional machine learning models, deep learning-based approaches are able to extract features autonomously from raw data. This allows for bypassing the feature engineering process, which is generally considered to be both error-prone and tedious. Moreover, deep learning strategies often outperform traditional models in terms of accuracy.