On the Decreasing Power of Kernel and Distance Based Nonparametric Hypothesis Tests in High Dimensions.
Aaditya RamdasSashank Jakkam ReddiBarnabás PóczosAarti SinghLarry A. WassermanPublished in: AAAI (2015)
Keyphrases
- high dimensions
- hypothesis tests
- high dimensional data
- kernel density estimation
- hypothesis testing
- high dimensional
- model selection
- hypothesis test
- sample size
- distance measure
- high dimensional spaces
- poor quality
- data sets
- markov chain
- maximum likelihood
- euclidean distance
- statistical model
- statistical tests
- dimensionality reduction
- likelihood ratio
- feature space
- similarity measure
- feature selection