CMU SoS Lablet Quarterly Executive Summary - October 2020

A. Fundamental Research
High level report of result or partial result that helped move security science forward-- In most cases it should point to a "hard problem". These are the most important research accomplishments of the Lablet in the previous quarter.

OVERALL SUMMARY

The CMU Science of Security Lablet has made advances in composability and usability within in security. This includes three technical thrusts: (1) contrastive explanation in automated planning to assist designers in understanding how agents made decisions in multi-objective planning; (2) evaluating leakage of training data from statistically-learned models when explainability is used to interpret model decisions; and (3) a field study to evaluate the effectiveness of a typestate language to verify blockchain programs used in developing countries. New this quarter are results from a summer program to train undergraduates in the Science of Security by introducing these students research experiences funded by the lablet. Most of these students came to CMU from universities that did not have active research programs, which broadens participation and diversity in research.