Impact of Data Pruning on Machine Learning Algorithm Performance

TL;DR

This study investigates how dataset pruning affects the performance of different machine learning algorithms in predicting movie ratings, revealing that the relative performance order remains consistent after pruning.

Contribution

It provides a comparative analysis of algorithm performance on pruned versus unpruned datasets using IMDb movie ratings, highlighting the impact of data pruning.

Findings

Better unpruned algorithm remains better after pruning

Pruning does not significantly change algorithm rankings

Dataset pruning improves model efficiency without altering performance hierarchy

Abstract

Dataset pruning is the process of removing sub-optimal tuples from a dataset to improve the learning of a machine learning model. In this paper, we compared the performance of different algorithms, first on an unpruned dataset and then on an iteratively pruned dataset. The goal was to understand whether an algorithm (say A) on an unpruned dataset performs better than another algorithm (say B), will algorithm B perform better on the pruned data or vice-versa. The dataset chosen for our analysis is a subset of the largest movie ratings database publicly available on the internet, IMDb [1]. The learning objective of the model was to predict the categorical rating of a movie among 5 bins: poor, average, good, very good, excellent. The results indicated that an algorithm that performed better on an unpruned dataset also performed better on a pruned dataset.