IFDS-MADLab Workshop

Statistical Approaches to Understanding Modern ML Methods

Aug 2-4, 2021
University of Wisconsin–Madison

When we use modern machine learning (ML) systems, the output often consists of a trained model with good performance on a test dataset. This satisfies some of our goals in performing data analysis, but leaves many unaddressed — for instance, we may want to build an understanding of the underlying phenomena, to provide uncertainty quantification about our conclusions, or to enforce constraints of safety, fairness, robustness, or privacy. As an example, classical statistical methods for quantifying a model’s variance rely on strong assumptions about the model — assumptions that can be difficult or impossible to verify for complex modern ML systems such as neural networks. 

This workshop will focus on using statistical methods to understand, characterize, and design ML models — for instance, methods that probe “black-box” ML models (with few to no assumptions) to assess their statistical properties, or tools for developing likelihood-free and simulation-based inference. Central themes of the workshop may include:

    • Using the output of a ML system to perform statistical inference, compute prediction intervals, or quantify measures of uncertainty

    • Using ML systems to test for conditional independence

    • Extracting interpretable information such as feature importance or causal relationships

    • Integrating likelihood-free inference with ML

    • Developing mechanisms for enforcing privacy, robustness, or stability constraints on the output of ML systems

    • Exploring connections to transfer learning and domain adaptation
    • Automated tuning of hyperparameters in black-box models and derivative-free optimization

More information at https://ifds.info/ifds-madlab-workshop/.

Two Faculty Spearhead New ML/AI Conferences

Po-Ling Loh

Po-Ling Loh was the one of the key people who created the Midwest Machine Learning Symposium. The symposium aims to convene regional machine learning researchers for stimulating discussions and debates, to foster cross-institutional collaboration, and to showcase the collective talent of machine learning researchers at all career stages. It will be held at the Logan Center at University of Chicago on June 6-7, 2018. The MMLS was founded in 2017 and Prof. Loh is this year’s chair. More information:

Dimitris Papailiopoulos

Dimitris Papailiopoulos co-chaired the first SysML Conference in Stanford, California February 15-16, 2018. SysML is a new conference targeting research at the intersection of systems and machine learning. The conference aimed to elicit new connections amongst these fields, including identifying best practices and design principles for learning systems, as well as developing novel learning methods and theory tailored to practical machine learning workflows. More information:   http://www.sysml.cc/2018/index.html


Computer-generated database of diffusion values is shared online

Dane Morgan
Dane Morgan
Materials Science & Engineering

University of Wisconsin—Madison engineers recently used powerful computers to quickly and accurately develop the world’s largest computed database of information about an important materials-mixing process called diffusion.

Led by Dane Morgan, Harvey D. Spangler Professor in Materials Science and Engineering at UW–Madison, the researchers published details of their advance July 19 in the journal Scientific Data. They also made the entire database freely available online, along with an online application to easily search and visualize the data and a utility called the Materials Simulation Toolkit (MAST) for engineers across the globe to access and use in their own materials design applications.

Read entire Scienceblog.com article