Machine Learning and Likelihood Free Inference in Particle Physics

[aaronjg]

https://figshare.com/articles/NIPS_2016_Keynote_Machine_Learning_Likelihood_Free_Inference_in_Particle_Physics/4291565/1

[darawk]

Anyone have a link to an actual paper on this? It seems interesting but this slideshow format is kind of hard to follow for me.

[cranmer → darawk]

Lots of topics and links throughout.

Approximate Bayesian Computation website does a good job of framing what is meant by likelihood free inference. https://approximatebayesiancomputational.wordpress.com/paper...

Here's an alternative technique for likelihood free inference: https://arxiv.org/abs/1506.02169 and a more recent approach http://beta.briefideas.org/ideas/5c2f74aedbf3618ca180382e393...

making machine learning more robust to systematic uncertainties https://arxiv.org/abs/1611.01046

A tech report summarizing Goodfellow's NIPS tutorial on GANs https://arxiv.org/abs/1701.00160

[jamessb → darawk]

It's an Keynote/invited talk, so there isn't a single corresponding paper as such. There are reference to papers on some of the slides:

* slide 75 gives a reference for CARL: https://arxiv.org/abs/1506.02169

* slides 93 gives 3 references for using deep learning to classify jet images https://arxiv.org/abs/1511.05190 https://arxiv.org/abs/1603.09349 https://www.arxiv.org/abs/1609.00607

* the reference for "Learning to Pivot with adversarial networks" is https://www.arxiv.org/abs/1611.01046