Clustering functional data using projections

来源 :上海交通大学 | 被引量 : 0次 | 上传用户：jbhjyh12345678

【摘要】

：

We show that,in the functional data context,by appropriately exploiting the functional nature of the data,it is possible to cluster the observations asymptotically perfectly in the case of different l

【作者】

：

Tung Huy Pham Aurore Delaigle Peter Hall

【机构】

：

University of Melbourne,3010,Australia

【出处】

：

上海交通大学

【发表日期】

：

2016年4期

【关键词】

：

asymptotically perfect clustering functional principal components Haar basis k-m

下载到本地 , 更方便阅读

下载此文赞助VIP

声明 : 本文档内容版权归属内容提供方 , 如果您对本文有版权争议 , 可与客服联系进行内容授权或下架

论文部分内容阅读

其他文献

Prediction risk for global-local shrinkage regression

Predictive performance in shrinkage regression suffers from two major difficulties:(ⅰ)the amount of relative shrinkage is monotone in the singular values of the design matrix and(ⅱ)the amount of shrin

会议

Error Bounds for Sequential Monte Carlo Samplers for Multimodal Distributions

In this talk,we provide bounds on the asymptotic variance for a class of sequential Monte Carlo(SMC)samplers designed for approximating multimodal distributions.

会议

SMC samplersasymptotic variancesharp peaksCurie-Weiss modelPotts model

REDESIGNING STATISTICS EDUCATION FOR A DATA-RICH WORLD

The uses of statistics in daily life and in the workforce are changing rapidly as data become more available to everyone.

会议

Statistics educationstatistics curriculumstatistical reasoningstatistical lit

Nonparametric Inference of Right Censored Data Using Smoothing Splines

This paper introduces a penalized nonparametric maximum likelihood estimation of the log-hazard function in analyzing right censored data.

会议

Functional Bahadur representationNonparametric inferencePenalized likelihoodR

Bayesian time series regression analysis with nonparametric modeling of autocorrelation

Time series regression has two purposes-comprehend the functional dependence of variable of interest on covariates and forecast the dependent variable for future values of covariates.

会议

Autocorrelation functionBayesian frameworkNonparametric modelingWhittle likel

POWER OF OPTIMAL SIGNAL DETECTION METHODS UNDER FINITE SAMPLE

In big data analysis for detecting rare and weak signals among n features,the Higher Criticism test(HC),Berk-Jones test(B-J),and some (o) -divergence tests have been proven optimal under the asymptoti

会议

Higher criticismphi-divergencegoodness-of-fitsparse mixturesignal detection

OPTIMAL BAYESIAN MINIMAX RATE FOR UNCONSTRAINED LARGE COVARIANCE MATRICES

There have been many researches on the minimax convergence rates for covariance matrices under various covariance classes.

会议

Minimax rateExpected posterior riskDecision theoretic prior selectionCovarian

Shrinkage-based diagonal Hotelling's tests for high-dimensional small sample size data

High-throughput expression profiling techniques bring novel tools and also statis-tical challenges to genetic research.

会议

Diagonal Hotelling's testHigh-dimensional dataMicroarray dataNull distri

Diagnostic analysis for Feller-Pareto linear regression model

In this paper,we introduce a Feller-Pareto linear regression(FPLR)model based on the Feller-Pareto distribution.

会议

Feller-pareto distributionFeller-Pareto linear regression modelCase-deletion m

On the double bootstrap for estimating the distribution of descriptive statistics of functional data

We present a double bootstrap procedure for reducing coverage error in confidence intervals of summary statistics for independent and identically distributed functional data.

会议

confidence interval calibrationcoverage erroriterated bootstrapresampling

Clustering functional data using projections

与本文相关的学术论文