Skip to main content

Table 2 Simulation results for model 1

From: Sparse sliced inverse regression for high dimensional data analysis

Model 1

(\(\rho\), p)

 

CISESIR

CISELDA

MGSDA

PLDA

MSDA

(0.5, 50)

\(\Delta\)

0.761

0.821

1.346

0.251

1.328

MSR

0.128

0.127

0.137

0.125

0.137

TPR

0.949

0.760

0.775

1.000

0.860

FPR

0.100

0.227

0.101

0.225

0.264

(0.5, 500)

\(\Delta\)

0.797

0.888

1.374

0.455

1.315

MSR

0.132

0.128

0.139

0.125

0.138

TPR

0.897

0.985

0.733

1.000

0.810

FPR

0.052

0.076

0.010

0.011

0.011

(0.5, 1000)

\(\Delta\)

0.632

0.604

1.384

0.406

1.307

MSR

0.129

0.125

0.139

0.129

0.135

TPR

0.932

0.999

0.739

1.000

0.794

FPR

0.026

0.066

0.007

0.176

0.005

(0.9, 50)

\(\Delta\)

1.070

1.031

1.714

0.140

1.672

MSR

0.209

0.207

0.213

0.206

0.215

TPR

0.835

0.925

0.368

1.000

0.481

FPR

0.140

0.214

0.037

0.254

0.164

(0.9, 500)

\(\Delta\)

0.925

1.086

1.730

0.409

1.703

MSR

0.216

0.215

0.217

0.209

0.213

TPR

0.828

0.998

0.376

1.000

0.399

FPR

0.067

0.160

0.015

0.401

0.007

(0.9, 1000)

\(\Delta\)

0.588

0.663

1.047

0.289

1.680

MSR

0.210

0.209

0.214

0.206

0.213

TPR

0.942

1.000

0.393

1.000

0.427

FPR

0.056

0.072

0.004

0.153

0.005

  1. \(\Delta\) is as defined in (10); TPR is the true positive rate; FPR is the false positive rate; MSR is the misclassification rate over a test set of 900 observations. Note again, TPR and FPR are with respect to variable selection. The reported numbers are averages over 50 repetitions