Quantifying Inherent Randomness in Machine Learning Algorithms

TL;DR

This study empirically quantifies how stochastic elements in training and data partitioning affect the performance variability of ML algorithms like RFs, GBMs, and FFNNs, highlighting the importance of controlling randomness for reproducibility.

Contribution

It systematically compares the impact of training randomness and data splitting on performance variation across multiple ML algorithms.

Findings

FFNNs exhibit larger variation due to training randomness.

Data splitting causes more variation than training randomness.

Heterogeneous datasets amplify the impact of data splitting variability.

Abstract

Most machine learning (ML) algorithms have several stochastic elements, and their performances are affected by these sources of randomness. This paper uses an empirical study to systematically examine the effects of two sources: randomness in model training and randomness in the partitioning of a dataset into training and test subsets. We quantify and compare the magnitude of the variation in predictive performance for the following ML algorithms: Random Forests (RFs), Gradient Boosting Machines (GBMs), and Feedforward Neural Networks (FFNNs). Among the different algorithms, randomness in model training causes larger variation for FFNNs compared to tree-based methods. This is to be expected as FFNNs have more stochastic elements that are part of their model initialization and training. We also found that random splitting of datasets leads to higher variation compared to the inherent randomness from model training. The variation from data splitting can be a major issue if the original dataset has considerable heterogeneity. Keywords: Model Training, Reproducibility, Variation