Abraham Wald and his work on WWII bomber damage is a really cool story highlighting why we need to always challenge the way we look at data and consider what our data is really telling us, and what data we may be missing.
Tag: stuff from the interweb
If you are like me, you remember the days from your childhood of free ranging far from home for hours on end without your parents really knowing where you were. Whether that was walking a mile to the ballpark for baseball practice, or taking a skiff a few miles from home to go out fishing in the bay, my parents rarely knew where I was most of the day. That type of parenting today would have you locked up.
This is a Q&A with a pair of researchers who conducted an interesting study on how moral judgement impacts policy / lawmaking surrounding the regulation of parenting. It’s a very in depth look at the class implications of these new regulations; an interesting blend of behavioral sciences and perceived versus actual risk.
A humorous yet low tech and possibly effective method for warding off lion attacks on cattle herds in Africa – painting eyes on the rear ends of cows. The study is in it’s initial phases, but it looks promising. It also illustrates that the most simple of solutions can be the most effective; something we need to consider in our analytics methodologies.
The full article can be read here:
http://www.australiangeographic.com.au/news/2016/07/new-eye-opening-solution-to-scare-off-predators
This was a really beautifully done article on Sumo wrestling with some very well constructed infographics that followed Tufte’s principles of information design quite closely. Here’s an example of one of the graphics:
You can read the full article with more graphics here:
http://fivethirtyeight.com/features/the-sumo-matchup-centuries-in-the-making/
As a performance data scientist, my day job is about finding non-obvious data access patterns in workloads. These patterns can be leveraged to tune a system, or learn more about the behaviors of users driving these workloads. We can often tell a lot from the metadata without seeing the contents of the transactional data which may contain private or sensitive data. This leads us to develop broad transitional profiling methodologies that allow us to provide feedback loops into applications to self-tune configurations to optimize the cost of running a workload, or in some cases, provide insights back to operators about their users and their usage patterns.
Continue readingA really cool decision tree from NASA’s Apollo program. It gets you thinking about how to incorporate well vetted, seemingly forgotten process flows into current procedures like automated cluster monitoring and alerting.
This is a great example (in between the cussing) that highlights the danger of entity linking rules that glue people together based on limited data, or in some cases, a single attribute, like SSN.
This was a recent highly publicized event here are 4 year old kid in Egypt was sentenced to life in prison because of, well, really crappy entity resolution:
http://www.bbc.com/news/world-middle-east-35633314
This issue was essentially caused by a match on name (and not even exact), while no other consideration was given to other attributes that would have clearly shown the child that was sentenced was not the actual target. Here are the names referenced in the BBC report in the link I embedded above:
Continue readingThis was an interesting podcast from Dr. Art Markman and Dr. Bob Duke from UT Austin – they are featured on a weekly podcast on KUT Austin titled Two Guys on Your Head that typically covers really interesting topics. This podcast focused on low information decision making and how people with expertise in certain fields can rely on a variety of experiences to process many variables in reaching a decision. Most of this decision making is difficult for a person to verbalize, and even more challenging to replicate programmatically.
Interesting stuff. Have a listen:
[soundcloud]https://soundcloud.com/kutnews/two-guys-on-your-head-low-information-decision-making[/soundcloud]
A great piece from KUT’s Two Guys on Your Head on why failure is important in education. Our propensity for avoiding mistakes, and teaching our children that failure is not ok, leads to a lack of training in experimentation. There’s a great analogy in this bit on the use of training wheels on a child’s bike to allow for limited failure (tipping from side to side) while preventing catastrophic failure (falling over).
[soundcloud]https://soundcloud.com/kutnews/tgoyh101813[/soundcloud]