Methods for Quantifying Dataset Similarity: a Review, Taxonomy and Comparison

Abstract

Quantifying the similarity between datasets has widespread applications in statistics and machine learning. The performance of a predictive model on novel datasets, referred to as generalizability, depends on how similar the training and evaluation datasets are. Exploiting or transferring insights between similar datasets is a key aspect of meta-learning and transfer-learning. In simulation studies, the similarity between distributions of simulated datasets and real datasets, for which the performance of methods is assessed, is crucial. In two- or $k$-sample testing, it is checked, whether the underlying distributions of two or more datasets coincide. Extremely many approaches for quantifying dataset similarity have been proposed in the literature. We examine more than 100 methods and provide a taxonomy, classifying them into ten classes. In an extensive review of these methods the main underlying ideas, formal definitions, and important properties are introduced. We compare the 118 methods in terms of their applicability, interpretability, and theoretical properties, in order to provide recommendations for selecting an appropriate dataset similarity measure based on the specific goal of the dataset comparison and on the properties of the datasets at hand. An online tool facilitates the choice of the appropriate dataset similarity measure.

124 pages

References in corpus (10)

• DISCO analysis: A nonparametric extension of analysis of variance
• Learning Generative Models with Sinkhorn Divergences
• Geometric Dataset Distances via Optimal Transport
• FastMMD: Ensemble of Circular Discrepancy for Efficient Two-Sample Test
• Quadratic distances on probabilities: A unified foundation
• Distances between Data Sets Based on Summary Statistics
• An Efficient and Distribution-Free Two-Sample Test Based on Energy Statistics and Random Projections
• A new test for the multivariate two-sample problem based on the concept of minimum energy
• New graph-based multi-sample tests for high-dimensional and non-Euclidean data
• A new ranking scheme for modern data and its application to two-sample hypothesis testing

Cited by in corpus (1)

• Time-Series Forecasting in Smart Manufacturing Systems: An Experimental Evaluation of the State-of-the-art Algorithms