Four short links: 30 May 2016
Algorithmic Transparency, Replay Tests, Hack in the Box Presentations, and Human Perception
- Algorithmic Transparency via Quantitative Input Influence: Theory and Experiments with Learning Systems (PDF) — Our empirical validation with standard machine learning algorithms demonstrates that QII measures are a useful transparency mechanism when black box access to the learning system is available. In particular, they provide better explanations than standard associative measures for a host of scenarios that we consider. Further, we show that in the situations we consider, QII is efficiently approximable and can be made differentially private while preserving accuracy. (via CMU)
- Why do Record/Replay Tests of Web Applications Break? (Paper a Day) — Locators caused over 73% of the test breakages we observed, and attribute-based locators caused the majority of these.
- Hack in the Box 2016 Presentation PDFs — be sure to check out Creating your own bad USB device.
- 39 Studies About Human Perception in 30 Minutes — good bootcamp in science of perception, useful for those working in UI and visualisation areas.