Making Machine Learning Models Safer

Speaker: Dr Kowshik Thopalli, Lawrence Livermore National Laboratory

Title:  Making Machine Learning Models Safer: Data and Model Perspectives

Date: Wed, Nov 29, 4:00-5:15pm, ENGR 231

Abstract: As machine learning systems are increasingly deployed in real-world settings like healthcare, finance, and scientific applications, ensuring their safety and reliability is crucial. However, many state-of-the-art ML models still suffer from issues like poor out-of-distribution generalization, sensitivity to input corruptions, requiring large amounts of data, and inadequate calibration - limiting their robustness and trustworthiness for critical real-world applications.  In this talk, I will first present a broad overview of different safety considerations for modern ML systems. I will then proceed to discuss our recent efforts in making ML models safer from two complementary perspectives - (i) manipulating data and (ii) enriching the model capabilities by developing novel training mechanisms.  I will discuss our work on designing new data augmentation techniques for object detection followed by demonstrating how, in the absence of data from desired target domains of interest, one could leverage pre-trained generative models for efficient synthetic data generation.  Next, I will present a new paradigm of training deep networks called model anchoring and show how one could achieve similar properties to an ensemble but through a single model. I will specifically discuss how model anchoring can significantly enrich the class of hypothesis functions being sampled and demonstrate its effectiveness through its improved performance on several safety benchmarks. I will conclude by highlighting exciting future research directions for producing robust ML models through leveraging multi-modal foundation models.

Bio: Kowshik Thopalli is a Machine Learning Scientist and a post-doctoral researcher at Lawrence Livermore National Laboratory. His research focuses on developing reliable machine learning models that are robust under distribution shifts. He has published papers on a variety of techniques to address model robustness, including domain adaptation, domain generalization, and test-time adaptation using geometric and meta-learning approaches. His expertise also encompasses integrating diverse knowledge sources, such as domain expert guidance and generative models, to improve model data efficiency, accuracy, and resilience to distribution shifts.  He received his Ph.D. in 2023 from Arizona State University.