QuantifyML: How Good is my Machine Learning Model?

TL;DR

QuantifyML provides a precise method to evaluate how well machine learning models have learned and generalized, going beyond accuracy metrics by analyzing model behavior through formal verification and counting techniques.

Contribution

It introduces a novel approach that translates models into C programs and uses model checking and counting to quantify learnability, robustness, and safety.

Findings

Enables precise evaluation of model learnability and generalization.

Allows comparison of different machine learning algorithms.

Quantifies safety and robustness of models.

Abstract

The efficacy of machine learning models is typically determined by computing their accuracy on test data sets. However, this may often be misleading, since the test data may not be representative of the problem that is being studied. With QuantifyML we aim to precisely quantify the extent to which machine learning models have learned and generalized from the given data. Given a trained model, QuantifyML translates it into a C program and feeds it to the CBMC model checker to produce a formula in Conjunctive Normal Form (CNF). The formula is analyzed with off-the-shelf model counters to obtain precise counts with respect to different model behavior. QuantifyML enables i) evaluating learnability by comparing the counts for the outputs to ground truth, expressed as logical predicates, ii) comparing the performance of models built with different machine learning algorithms (decision-trees vs. neural networks), and iii) quantifying the safety and robustness of models.