New data analysis competitions
- Bimbo has a competition on Kaggle, with a monetary prize attached.
The Royal Canadian Mounted Police has been using mass surveillance devices known as IMSI catchers, in public, for a decade. In that time, the police have indiscriminately surveilled potentially thousands of Canadians without their knowledge, and stored that information for later use.
Our faces are readily accessible to other people, and most people must expose their faces to other people in order to participate in society. When we do so, there is very little that we can do as individuals to prevent other people from capturing the images of our faces and subjecting us to facial recognition technologies.
If someone stalks us or commits identify theft against us by using our passwords or credit card numbers, we can defend ourselves by simply changing those unique identifiers. We can even change our names. But contrary to what action movies suggest, we cannot change our faces.
A secret report warned that British spies may have put lives at risk because their surveillance systems were sweeping up more data than could be analyzed, leading them to miss clues to possible security threats.
- Safely Interruptible Agents (PDF, academic paper). From the abstract:
This paper explores a way to make sure a learning agent will not learn to prevent (or seek!) being interrupted by the environment or a human operator.
Effectively, a kill-switch.
R Passes SAS in Scholarly Use (finally). I had to use SAS in the past. I think I got PTSD due to that. Some other day I can talk about my SPSS experiences.
In a paper published Tuesday in the Journal of Oncology Practice, the trio detailed how they used anonymized Bing search logs to identify people whose queries provided strong evidence that they had recently been diagnosed with pancreatic cancer – a particularly deadly and fast-spreading cancer that is frequently caught too late to cure. Then they retroactively analyzed searches for symptoms of the disease over many months prior to identify patterns of queries most likely to signal an eventual diagnosis.
A Brexit prediction model: Understanding statistics through interactive visualizations.
A discussion on the subject: on whether y-axis labels are always necessary.