FastAMI -- a Monte Carlo Approach to the Adjustment for Chance in Clustering Comparison Metrics
This work addresses a computational bottleneck for researchers and practitioners comparing clusterings in machine learning, though it is incremental as it builds on existing AMI methods.
The authors tackled the computational difficulty of adjusting clustering comparison metrics for chance in large datasets by proposing FastAMI, a Monte Carlo-based method that approximates Adjusted Mutual Information (AMI) and extends it to Standardized Mutual Information (SMI). Their method is fast enough for large datasets and maintains more accurate results than a pairwise approach.
Clustering is at the very core of machine learning, and its applications proliferate with the increasing availability of data. However, as datasets grow, comparing clusterings with an adjustment for chance becomes computationally difficult, preventing unbiased ground-truth comparisons and solution selection. We propose FastAMI, a Monte Carlo-based method to efficiently approximate the Adjusted Mutual Information (AMI) and extend it to the Standardized Mutual Information (SMI). The approach is compared with the exact calculation and a recently developed variant of the AMI based on pairwise permutations, using both synthetic and real data. In contrast to the exact calculation our method is fast enough to enable these adjusted information-theoretic comparisons for large datasets while maintaining considerably more accurate results than the pairwise approach.