### ### Articles in ML/AI/StatLearning ### (updated 2020-12-08) ### ################################################################ Training Issue: Deepmind Protein(1) https://www.nature.com/articles/d41586-020-03348-4 Scientist not impressed(2) https://futurism.com/the-byte/scientists-unimpressed-googles-protein-folding-algorithm Hiden Weakness in ML https://www.discovermagazine.com/technology/google-reveals-major-hidden-weakness-in-machine-learning Way we train AI is flawed (Underspecification) https://www.technologyreview.com/2020/11/18/1012234/training-machine-learning-broken-real-world-heath-nlp-computer-vision/ Bias Issue: Bias issue in AI https://www.wsj.com/articles/how-to-make-artificial-intelligence-less-biased-11604415654?mod=flipboard AI in healthcare is racist https://www.zdnet.com/article/artificial-intelligence-in-healthcare-is-racist/ Training Demo: Article: Drive with NN https://hackaday.com/2020/11/07/training-a-neural-network-to-play-a-driving-game/ Uber-AI on Atari Games https://www.engadget.com/uber-ai-masters-atari-games-012832222.html Applications of ML: Motion Capture with only 1 camera w ML https://80.lv/articles/animation-deep-learning-vs-traditional-methods/ Automa te 3D-animation https://www.reddit.com/r/MachineLearning/comments/jps5uw/n_digital_domains_deformation_simulation_system/ Auto color BW pics: https://www.reddit.com/r/computervision/comments/jpqn86/this_ai_can_colorize_your_black_white_photos_with/ Archeological Site searched by ML https://www.nytimes.com/2020/11/24/science/artificial-intelligence-archaeology-cnn.html Rutgers researchers develop automated blood drawing robot https://www.dailytargum.com/article/2020/02/rutgers-researchers-develop-automated-blood-drawing-robot Find Pique replacemwnt with clustering: https://cestlaz.github.io/stories/emacs/ Language Processing: https://syncedreview.com/2020/11/09/amazon-alexa-ais-language-model-is-all-you-need-explores-nlu-as-qa/ GAN: Chinese Painting https://syncedreview.com/2020/11/30/princeton-students-ai-model-generates-chinese-landscape-paintings-that-fool-human-evaluators/