03543nam 2201009z- 450 991057687320332120220621(CKB)5720000000008444(oapen)https://directory.doabooks.org/handle/20.500.12854/84584(oapen)doab84584(EXLCZ)99572000000000844420202206d2022 |y 0engurmn|---annantxtrdacontentcrdamediacrrdacarrierNonparametric Statistical Inference with an Emphasis on Information-Theoretic MethodsMDPI - Multidisciplinary Digital Publishing Institute20221 online resource (226 p.)3-0365-4297-3 3-0365-4298-1 This book addresses contemporary statistical inference issues when no or minimal assumptions on the nature of studied phenomenon are imposed. Information theory methods play an important role in such scenarios. The approaches discussed include various high-dimensional regression problems, time series and dependence analyses.History of engineering and technologybicsscMechanical engineering and materialsbicsscTechnology: general issuesbicsscadaptive splinesarchimedean copulaB-splineschange pointsCIFECMIconditional infomax feature extractionconditional mutual informationconsistencyconsistent selectionentropyestimationextreme-value copulagaussian mixturegeneralized information criteriongenerative tree modelhigh-dimensional regressionhigh-dimensional time seriesinfluence functionsinformation measuresinformation theoryJMIjoint mutual information criterionkernel estimationlearning systemsloss functionMarkov blanketmaximum likelihood estimationminimum distance estimationmisclassification riskmisspecificationmodel misspecificationmultivariate analysisn/anetwork estimationnonparametric variable selection criterianonstationarityparameter estimationpenalized estimationprediction methodsprivacyrandom predictorsright-censored datarobustnesssemiparametric regressionstatistical learning theorysubgaussianitysupervised classificationsynthetic data transformationtail dependencytime seriesvariable selection consistencyHistory of engineering and technologyMechanical engineering and materialsTechnology: general issuesMielniczuk Janedt1325229Mielniczuk JanothBOOK9910576873203321Nonparametric Statistical Inference with an Emphasis on Information-Theoretic Methods3036710UNINA