A few hand-picked company names from the Panama Papers database
I'm back on the Monistic Almanac project. Doing some fruitful research:
📰 Write up of my recent workshop at the MA Graphic Media Design, LCC.
Before it completely dies in darkness, here is a proper ending to this log-thread.
The project is now online at: http://countingthefuture.net/
and the code: https://github.com/davidbenque/counting-the-future
This will keep growing for the foreseeable future 😉
- - - - - - [ END OF THREAD ] - - - - - -
Bonus: great physical model of iris features in Anderson (1936), he actually gathered the iris data and shared it with Fisher.
The citation link between Fisher's linear separation of the Iris dataset (Annals of Eugenics, 1936)—widely used as a demo and/or benchmark for classifiers—and Support-Vector Machines from Cortes and Vapnik (1995).
more on SVMs in Mackenzie's Machines Learners (ch.6 Patterns and Differences)
Cortes and Vapnik: https://link.springer.com/content/pdf/10.1007/BF00994018.pdf
On the development front, made some progress this week. The graph view is taking shape (although plenty of 'special cases' to hard code which is part of the fun), and image view + reference view were nice and easy to implement.
the back end is getting ... rhizomatic
#flask #d3js #neo4j
https://mastodon.social/media/Ucrw75guO6tJm9AUvxs https://mastodon.social/media/4ovkwWAwKRKFof6aM2w https://mastodon.social/media/hMu5Gq8OYZazOxOm8R0 https://mastodon.social/media/xACScVo55ocjfQJNMPk
The "phrenology" line—predicting criminality from facial features—going from 1876 and Lombroso's Criminal Man, straight to ... 2016 and ‘Automated Inference on Criminality using Face Images’
📰 I finally posted my talk from Anticipation2017 about almanacs
a quick thread of Bloomberg terminals, the $1500 a month oracle.
Sourced from this AIGA article:
and this timeline of keyboards:
A couple related papers:
Terminal value: Building the alternative Bloomberg
From Pipes to Scopes: The Flow Architecture of Financial Markets
new word: 🔭 scopic 🔭
Extending the 'Social Physics' thread from earlier to its current incarnation. First as Alex Pentland's Human Dynamics Lab at MIT, then with the inevitable spinoff startup, complete with a crypto-currency (and upcoming ICO) for "unlimited accurate predictions".
Carr's critical review: https://www.technologyreview.com/s/526561/the-limits-of-social-engineering/
Mattern, S. (2015) ‘Mission Control: A History of the Urban Dashboard’, Places Journal. https://placesjournal.org/article/mission-control-a-history-of-the-urban-dashboard/
🎥 Find of the day: *Data for Decision*, a 1968 short film by David Millar and the National Film Board of Canada.
Revisited a key paper today, 'Big Data, social physics, and spatial analysis: The early years', Barnes & Wilson, 2014. http://journals.sagepub.com/doi/full/10.1177/2053951714535365
Mapping out the trajectory from Quetelet to GIS at Harvard put the editor to the test. It works! much easier than anything I was able to do with CSVs, and more flexible. Not sure this if this is truly rhizomatic as Corner or Deleuze would have it, but it definitely feels like 'burrowing'.
Prior art: Histograph is a digital humanities project mapping relationships between historical figures based on co-occurrences of names in documents. It also uses a #neo4j database, although in the 'network changing over time' way, where I have gone with a 'the network is the timeline' approach.
👍 Wired: history, graph database, european integration.
👎 Tired: big data, very hard to use, comical computational mistakes.