Bayesian Performance Analysis for Algorithm Ranking Comparison

• Published in: IEEE transactions on evolutionary computation Vol. 26; no. 6; p. 1
• Main Authors: Rojas-Delgado, Jairo, Ceberio, Josu, Calvo, Borja, Lozano, Jose A.
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.12.2022; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Bayes methods; Bayesian analysis; Bayesian inference; benchmarking; Data models; evolutionary algorithms; Evolutionary computation; Inference algorithms; Machine learning; Optimization; Performance assessment; Permutations; probabilistic models on permutation spaces; Ranking; Ratings & rankings; Sociology; Source code; Statistical analysis; Statistical tests; Statistics; Uncertainty
• ISSN: 1089-778X; 1941-0026; 1941-0026
• DOI: 10.1109/TEVC.2022.3208110

In the field of optimization and machine learning, the statistical assessment of results has played a key role in conducting algorithmic performance comparisons. Classically, null hypothesis statistical tests have been used. However, recently, alternatives based on Bayesian statistics have shown great potential in complex scenarios, especially when quantifying the uncertainty in the comparison. In this work, we delve deep into the Bayesian statistical assessment of experimental results by proposing a framework for the analysis of several algorithms on several problems/instances. To this end, experimental results are transformed to their corresponding rankings of algorithms, assuming that these rankings have been generated by a probability distribution (defined on permutation spaces). From the set of rankings, we estimate the posterior distribution of the parameters of the studied probability models, and several inferences concerning the analysis of the results are examined. Particularly, we study questions related to the probability of having one algorithm in the first position of the ranking or the probability that two algorithms are in the same relative position in the ranking. Not limited to that, the assumptions, strengths, and weaknesses of the models in each case are studied. To help other researchers to make use of this kind of analysis, we provide a Python package and source code implementation at.