Pinboard (cshalizi) https://pinboard.in/u:cshalizi/public/ recent bookmarks from cshalizi [2502.20755] Minimax Optimal Kernel Two-Sample Tests with Random Features 2025-03-16T19:31:58+00:00 https://arxiv.org/abs/2502.20755 cshalizi to:NB hilbert_space statistics two-sample_tests random_features https://pinboard.in/ https://pinboard.in/u:cshalizi/b:e765167ec668/ Likelihood-Free Hypothesis Testing and Applications of the Energy Distance 2024-12-06T14:04:25+00:00 https://dspace.mit.edu/handle/1721.1/155358 cshalizi to:NB to_read hypothesis_testing two-sample_tests statistics via:_onionesque kernel_methods goodness-of-fit https://pinboard.in/ https://pinboard.in/u:cshalizi/b:d56e7c266c6a/ [2109.03582] Higher Order Kernel Mean Embeddings to Capture Filtrations of Stochastic Processes 2023-02-15T19:57:50+00:00 https://arxiv.org/abs/2109.03582 cshalizi to:NB stochastic_processes kernel_methods causal_discovery time_series statistical_inference_for_stochastic_processes hilbert_space re:codename:catherine_wheel two-sample_tests statistics have_read path_signatures https://pinboard.in/ https://pinboard.in/u:cshalizi/b:105636bb7bad/ [2105.03481] Stein's Method Meets Statistics: A Review of Some Recent Developments 2021-05-12T18:13:21+00:00 https://arxiv.org/abs/2105.03481 cshalizi to:NB steins_method probability statistics monte_carlo two-sample_tests goodness-of-fit re:codename:catherine_wheel https://pinboard.in/ https://pinboard.in/u:cshalizi/b:42b841f98d00/ Kim , Ramdas , Singh , Wasserman : Classification accuracy as a proxy for two-sample testing 2021-02-04T15:31:39+00:00 https://projecteuclid.org/euclid.aos/1611889233 cshalizi0ϵ>0 as d,n→∞d,n→∞, then (a) the permutation-based test is consistent (has power approaching to one), (b) a computationally efficient test based on a Gaussian approximation of the null distribution is also consistent. To get a finer understanding of the rates of consistency, we study a specialized setting of distinguishing Gaussians with mean-difference δδ and common (known or unknown) covariance ΣΣ, when d/n→c∈(0,∞)d/n→c∈(0,∞). We study variants of Fisher’s linear discriminant analysis (LDA) such as “naive Bayes” in a nontrivial regime when ϵ→0ϵ→0 (the Bayes classifier has true accuracy approaching 1/2), and contrast their power with corresponding variants of Hotelling’s test. Surprisingly, the expressions for their power match exactly in terms of nn, dd, δδ, ΣΣ, and the LDA approach is only worse by a constant factor, achieving an asymptotic relative efficiency (ARE) of 1/π‾‾√1/π for balanced samples. We also extend our results to high-dimensional elliptical distributions with finite kurtosis. Other results of independent interest include minimax lower bounds, and the optimality of Hotelling’s test when d=o(n)d=o(n). Simulation results validate our theory, and we present practical takeaway messages along with natural open problems."]]> to:NB hypothesis_testing two-sample_tests classifiers high-dimensional_statistics heard_the_talk kith_and_kin singh.aarti wasserman.larry ramdas.aaditya https://pinboard.in/ https://pinboard.in/u:cshalizi/b:9a8de542290c/ [2012.09828] Nonparametric Two-Sample Hypothesis Testing for Random Graphs with Negative and Repeated Eigenvalues 2020-12-18T10:33:02+00:00 https://arxiv.org/abs/2012.09828 cshalizi to:NB network_data_analysis re:network_differences two-sample_tests https://pinboard.in/ https://pinboard.in/u:cshalizi/b:aa95516da3d0/ [1506.02785] On the Error of Random Fourier Features 2020-12-14T01:48:11+00:00 https://arxiv.org/abs/1506.02785 cshalizi random_features kernel_methods approximation computational_statistics concentration_of_measure two-sample_tests regression schneider.jeff have_read to_teach:childs_garden_of_statistical_learning_theory in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:bdadd4e91fb9/ [1903.11117] Testing for Differences in Stochastic Network Structure 2020-11-25T14:52:28+00:00 https://arxiv.org/abs/1903.11117 cshalizi to:NB network_data_analysis re:network_differences two-sample_tests hypothesis_testing to_read https://pinboard.in/ https://pinboard.in/u:cshalizi/b:ba62b7e5b0dc/ Ghoshdastidar , Gutzeit , Carpentier , von Luxburg : Two-sample hypothesis testing for inhomogeneous random graphs 2020-11-18T21:44:11+00:00 https://projecteuclid.org/euclid.aos/1597370670 cshaliziρd(P,Q)>ρ. We observe that if mm is small, then the minimax separation is too large for some popular choices of dd, including total variation distance between corresponding distributions. This implies that some models that are widely separated in dd cannot be distinguished for small mm, and hence, the testing problem is generally not solvable in these cases. "We also show that if m>1m>1, then the minimax separation is relatively small if dd is the Frobenius norm or operator norm distance between PP and QQ. For m=1m=1, only the latter distance provides small minimax separation. Thus, for these distances, the problem is solvable for small mm. We also present near-optimal two-sample tests in both cases, where tests are adaptive with respect to sparsity level of the graphs."]]> to:NB to_read statistics two-sample_tests network_data_analysis re:network_differences https://pinboard.in/ https://pinboard.in/u:cshalizi/b:979b239f43f9/ Kim , Lee , Lei : Global and local two-sample tests via regression 2020-11-16T16:11:48+00:00 https://projecteuclid.org/euclid.ejs/1576573369 cshalizi to:NB two-sample_tests nonparametrics high-dimensional_statistics regression kith_and_kin lee.ann_b. lei.jing heard_the_talk https://pinboard.in/ https://pinboard.in/u:cshalizi/b:d95b8656b5cd/ [1810.11953] Failing Loudly: An Empirical Study of Methods for Detecting Dataset Shift 2019-10-29T15:00:54+00:00 https://arxiv.org/abs/1810.11953 cshalizi to:NB dataset_shift machine_learning model_checking lipton.zachary two-sample_tests statistics https://pinboard.in/ https://pinboard.in/u:cshalizi/b:7ff5939f686e/ [1910.08883] The Exact Equivalence of Independence Testing and Two-Sample Testing 2019-10-22T13:44:52+00:00 https://arxiv.org/abs/1910.08883 cshalizi to:NB two-sample_tests dependence_measures statistics hypothesis_testing https://pinboard.in/ https://pinboard.in/u:cshalizi/b:348b88959e6d/ [1909.13464] Network Differential Connectivity Analysis 2019-10-01T16:17:41+00:00 https://arxiv.org/abs/1909.13464 cshalizi to:NB network_data_analysis hypothesis_testing two-sample_tests statistics re:network_differences https://pinboard.in/ https://pinboard.in/u:cshalizi/b:612fcc0d1d37/ [1602.02210] Classification accuracy as a proxy for two sample testing 2019-05-28T17:02:27+00:00 https://arxiv.org/abs/1602.02210 cshalizi to:NB classifiers two-sample_tests statistics hypothesis_testing kith_and_kin ramdas.aaditya wasserman.larry singh.aarti https://pinboard.in/ https://pinboard.in/u:cshalizi/b:8d9ae6124edc/ Training generative neural networks via maximum mean discrepancy optimization 2015-07-15T14:02:10+00:00 http://auai.org/uai2015/proceedings/papers/230.pdf cshalizi to:NB simulation stochastic_models neural_networks machine_learning two-sample_tests hypothesis_testing nonparametrics kernel_methods statistics computational_statistics ghahramani.zoubin https://pinboard.in/ https://pinboard.in/u:cshalizi/b:34338d71a393/ [1411.2045] Multivariate f-Divergence Estimation With Confidence 2015-01-22T05:04:09+00:00 http://arxiv.org/abs/1411.2045 cshalizi estimation entropy_estimation information_theory statistics two-sample_tests in_NB hero.alfred_o._iii https://pinboard.in/ https://pinboard.in/u:cshalizi/b:0dbac386b55d/ [1409.2344] A nonparametric two-sample hypothesis testing problem for random dot product graphs 2015-01-20T13:33:07+00:00 http://arxiv.org/abs/1409.2344 cshalizi network_data_analysis hypothesis_testing two-sample_tests re:network_differences statistics to_read in_NB to_teach:graphons https://pinboard.in/ https://pinboard.in/u:cshalizi/b:041f7a5d64d9/ [1407.1212] Comparison of multivariate distributions using quantile-quantile plots and related tests 2015-01-20T01:59:24+00:00 http://arxiv.org/abs/1407.1212 cshalizi goodness-of-fit two-sample_tests statistics in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:f635706b9701/ [1405.0558] The Falling Factorial Basis and Its Statistical Applications 2014-12-02T00:39:08+00:00 http://arxiv.org/abs/1405.0558 cshalizi to:NB have_read splines nonparametrics statistics two-sample_tests kith_and_kin tibshirani.ryan https://pinboard.in/ https://pinboard.in/u:cshalizi/b:dfcdfbd65639/ [1001.0591] Comparing Distributions and Shapes using the Kernel Distance 2014-10-16T15:15:28+00:00 http://arxiv.org/abs/1001.0591 cshalizi to:NB kernel_estimators two-sample_tests statistics probability re:network_differences have_read https://pinboard.in/ https://pinboard.in/u:cshalizi/b:4ffff8f257a3/ [1406.2083] Kernel MMD, the Median Heuristic and Distance Correlation in High Dimensions 2014-07-12T00:29:26+00:00 http://arxiv.org/abs/1406.2083 cshalizi to:NB hypothesis_testing two-sample_tests kernel_estimators dependence_measures kith_and_kin wasserman.larry singh.aarti ramdas.aaditya high-dimensional_statistics https://pinboard.in/ https://pinboard.in/u:cshalizi/b:ca409b80b289/ [1210.4584] Two-Sample Testing in High-Dimensional Models 2014-03-05T14:21:05+00:00 http://arxiv.org/abs/1210.4584 cshalizi hypothesis_testing high-dimensional_statistics two-sample_tests statistics re:network_differences in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:e310404d6b50/ [1307.1954] B-tests: Low Variance Kernel Two-Sample Tests 2014-02-11T21:35:29+00:00 http://arxiv.org/abs/1307.1954 cshalizi to:NB two-sample_tests kernel_methods hilbert_space hypothesis_testing statistics https://pinboard.in/ https://pinboard.in/u:cshalizi/b:94ff051ff32e/ [1207.6076] Equivalence of distance-based and RKHS-based statistics in hypothesis testing 2013-11-17T20:15:33+00:00 http://arxiv.org/abs/1207.6076 cshalizi kernel_methods hilbert_space two-sample_tests statistics nonparametrics to_read in_NB entableted independence_tests https://pinboard.in/ https://pinboard.in/u:cshalizi/b:4290f416a332/ [1305.0423] Testing Hypotheses by Regularized Maximum Mean Discrepancy 2013-05-03T14:57:03+00:00 http://arxiv.org/abs/1305.0423 cshalizi two-sample_tests statistics hilbert_space kernel_methods to_read in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:bca72578fa28/ [1304.4564] A high-dimensional two-sample test for the mean using random subspaces 2013-04-23T18:06:43+00:00 http://arxiv.org/abs/1304.4564 cshalizi random_projections two-sample_tests hypothesis_testing statistics in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:c1229ad954ac/ [1304.5939] Exact and asymptotically robust permutation tests 2013-04-23T18:01:18+00:00 http://arxiv.org/abs/1304.5939 cshalizi hypothesis_testing two-sample_tests statistics in_NB have_read https://pinboard.in/ https://pinboard.in/u:cshalizi/b:65e76d4ff0ca/ Relative Density-Ratio Estimation for Robust Distribution Comparison 2013-04-04T17:32:51+00:00 http://www.mitpressjournals.org/doi/abs/10.1162/NECO_a_00442 cshalizi density_estimation statistics two-sample_tests goodness-of-fit have_read in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:f476c5fd4154/ [1304.0796] Direction-Projection-Permutation for High Dimensional Hypothesis Tests 2013-04-04T16:35:06+00:00 http://arxiv.org/abs/1304.0796 cshalizi two-sample_tests high-dimensional_statistics goodness-of-fit hypothesis_testing statistics to_teach:undergrad-ADA have_read in_NB visual_display_of_quantitative_information https://pinboard.in/ https://pinboard.in/u:cshalizi/b:94595c32ca65/ Sriperumbudur , Fukumizu , Gretton , Schölkopf , Lanckriet : On the empirical estimation of integral probability metrics 2012-09-18T14:08:02+00:00 http://projecteuclid.org/DPubS?service=UI&version=1.0&verb=Display&handle=euclid.ejs/1347974672 cshalizi to_read kernel_methods statistics probability two-sample_tests in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:ca18d39b56f2/ Modern Two-Sample Tests « Normal Deviate 2012-07-14T15:50:36+00:00 http://normaldeviate.wordpress.com/2012/07/14/modern-two-sample-tests/ cshalizitwo-sample_tests statistics hypothesis_testing wasserman.larry https://pinboard.in/ https://pinboard.in/u:cshalizi/b:7d452ff9efd6/ A Kernel Two-Sample Test 2012-04-03T00:34:03+00:00 http://jmlr.csail.mit.edu/papers/v13/gretton12a.html cshalizi to_read hilbert_space kernel_methods goodness-of-fit statistics concentration_of_measure probability two-sample_tests re:network_differences in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:2958bc7b3490/ Henze : A Multivariate Two-Sample Test Based on the Number of Nearest Neighbor Type Coincidences 2012-02-17T22:53:30+00:00 http://projecteuclid.org/DPubS?service=UI&version=1.0&verb=Display&handle=euclid.aos/1176350835 cshalizi to_read statistics hypothesis_testing two-sample_tests re:AoS_project in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:bb7bbf941041/ [1202.1561] Tree Models for Difference and Change Detection in a Complex Environment 2012-02-10T05:18:16+00:00 http://arxiv.org/abs/1202.1561 cshalizi re:network_differences statistics hypothesis_testing density_estimation decision_trees have_read data_mining two-sample_tests in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:1d69327d5561/ Scheffe : Statistical Inference in the Non-Parametric Case (1943) 2012-02-08T21:30:25+00:00 http://projecteuclid.org/DPubS?service=UI&version=1.0&verb=Display&handle=euclid.aoms/1177731355 cshalizi have_read statistics nonparametrics history_of_statistics estimation hypothesis_testing two-sample_tests in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:e1e4a2fb000b/ f-Divergence Estimation and Two-Sample Homogeneity Test Under Semiparametric Density-Ratio Models 2012-02-07T19:15:38+00:00 http://ieeexplore.ieee.org/xpls/abs_all.jsp?isnumber=6145469&arnumber=6018305&tag=1 cshalizi statistics density_estimation information_theory hypothesis_testing two-sample_tests in_NB density_ratio_estimation https://pinboard.in/ https://pinboard.in/u:cshalizi/b:d50cd7cd174b/ Nonparametric Tests for Homogeneity Based on Non-Bipartite Matching 2012-01-16T16:13:17+00:00 http://pubs.amstat.org/doi/abs/10.1198/jasa.2011.tm10576 cshalizi statistics hypothesis_testing density_estimation change-point_problem two-sample_tests in_NB https://pinboard.in/ https://pinboard.in/u:cshalizi/b:53436bb49b5d/