emanuelhuber / BLP

Bayesian Learning Potential

Bayesian Learning Potential

bibliography: http://andrius.velykis.lt/2012/06/master-bibtex-file-git-submodules/

https://gist.github.com/201827a159120cb808ced9d43bc30ac6.git

Distances

Check:

• https://statistik.econ.kit.edu/download/doc_secure1/3_StochModels.pdf

How do the distances reacts? -> https://www.datadoghq.com/blog/engineering/robust-statistical-distances-for-machine-learning/#cramrvon-mises-distance

http://www.stat.cmu.edu/~larry/=sml/Opt.pdf

Kolmogorov distance

https://stackoverflow.com/a/54138339

1-Wasserstein distance

https://stats.stackexchange.com/a/299391

∫∞x=−∞|F(x)−G(x)|dx,

Check

http://zevross.com/blog/2017/06/19/tips-and-tricks-for-working-with-images-and-figures-in-r-markdown-documents/

• Distance between CDF
• convergence analysis
• Measure of reliability of the distance (p-value?)
• Clustering
• Basic example
• Hydrogeological example
• Multi-dimensional CDF - how to do?

Basic examples

https://www.sciencedirect.com/science/article/pii/S1364815215000237#fd9

Ishigami–Homma function

(Eq. (4.34) in Saltelli et al. (2008))

y = sin(x₁) + a sin(x₂)² + b x₃⁴ sin(x₁)

where all x_i follow a uniform distribution over [-pi, pi], and a = 2 and b = 1.