From JtkwikiJump to navigationJump to search
This page contains interesting stuff I came across at conferences.
Open Data Science Conference 2016
- reproducible computing in dathttp://pipeline.io/a science, by Douglas Ashton
- An intro into Single Layer ANNs and Gradient Descent
- 3 Challenges for Open Data Science by Neil Lawrence -- mostly stuff that I found relatively well known already, but the mention of the ratio of information processing to information transmission was very welcome to me.
- The Role of Constructivism in Data Science -- should there be one, really? Constructivism has two faces / interpretation, one is the outright denial of any objectivity and the other is a kind of acknowledgement of subjectivity. There should be no role for the former in any kind of science, but the latter may have some points.
- Chris Fregly gave a talk / demo of a Data Science pipeline comprised of a huge number of data-sciencey components. It was not clear to me whether the demo application really required any of these high performance / high throughput / highly scalable components, but it was surely interesting to see an example of tying all these together. See also  (I think they're related).
- Openscoring (incl. Openscoring github site)
- Dynomite (github site) adding functionality to distribute data structure storage / caching