Mikael Huss

Nice work from @infoecho on Convolutional Neural Networks for Genomic Variant Calling with @tensorflow:https://medium.com/towards-data-science/simple-convolution-neural-network-for-genomic-variant-calling-with-tensorflow-c085dbc2026f …

Simple Convolutional Neural Network for Genomic Variant Calling with TensorFlow. just write-up some thought:https://medium.com/@infoecho/simple-convolution-neural-network-for-genomic-variant-calling-with-tensorflow-c085dbc2026f …

A neural net tried to name paint colours & it didn't go well. Unless you want to repaint your house in "Stanky Bean"?http://lewisandquark.tumblr.com/post/160776374467/new-paint-colors-invented-by-neural-network …

Challenging some deep learning preconceptions (e.g. "DL is useless for low sample sizes")http://hyperparameter.space/blog/when-not-to-use-deep-learning/ …

Some nice contextualization of deep learning/AI for drug discovery here:https://hackernoon.com/make-pharma-great-again-with-artificial-intelligence-some-challenges-50e91ea9988d …

"Git LFS works like a charm when things are done perfectly and in a very precise way" but when they aren't, read on:http://vooban.com/en/tips-articles-geek-stuff/migrating-to-git-lfs-for-developing-deep-learning-applications-with-large-files/ …

ADAGE signature analysis: differential expression analysis with data-defined gene sets @GreeneScientisthttp://biorxiv.org/content/early/2017/06/27/156620 …

AI progress measurement by EFFhttps://www.eff.org/ai/metrics