Final project for CS230: Deep Learning (Robert Moss: mossr@cs.stanford.edu), consulting with Bernard Lange, who's taking CS330: Meta Learning.
When validating a black-box system, exhaustively evaluating over the entire validation dataset may be computationally intractable. The challenge then becomes to intelligently automate selective validation given knowledge of the system failures experienced so far. We propose an adaptive black-box validation framework that will learn system weaknesses over time and exploit this knowledge to propose validation samples that will likely result in a failure. We use a low-dimensional encoded representation of inputs to train an adversarial failure classifier to intelligently select candidate failures to evaluate. Experiments were run to test our approach against a random candidate selection process and we also compare against full knowledge of the true system failures. We stress test a black-box neural network classifier trained on the MNIST dataset. Results show that using our framework, the adversarial failure classifier selects failures about 3 times more often than random.
- Paper: Adversarial Weakness Recognition for Efficient Black-Box Validation
- Presentation:
Open julia from the root directory, go into the package mode using ] and run:
dev .then you can run
using WeaknessRecognitionRun the src/run.jl file to load the models and call the main sampled_validation_iteration loop.