NEXT: a software system prototyping interactive machine learning algorithms and applications
Chicago Crime demo: Network estimation with partially observed crime data
Our software has been running the The New Yorker Caption Contest (https://newyorker.com/cartoons/vote) since 2016, initially with the NEXT active machine learning system (http://nextml.org). Over 70 million human responses have been recorded while running these contests and are available at https://github.com/nextml/caption-contest-data.
Deep variational canonical correlation analysis: http://ttic.uchicago.edu/~wwang5/papers/vcca_tf0.9_code.tgz
Deep CCA and related methods: http://ttic.uchicago.edu/~klivescu/software/dcca.tgz