Skip to main content

Table 1 Comparisons of seven classification methods on simulated data

From: The application of sparse estimation of covariance matrix to quadratic discriminant analysis

Methods

ISSC

ISDC

DSSC

DSDC

DLDA

0.048 (0.015, 50)

0.083 (0.015, 100)

0.228 (0.02, 1025)

0.217 (0.04, 1175)

DQDA

0.049 (0.021, 50)

0.013 (0.007, 50)

0.243 (0.025, 1400)

0.214 (0.032, 825)

NN

0.056 (0.02, 100)

0.424 (0.021, 50)

0.27 (0.034, 575)

0.112 (0.061, 475)

SVM

0.054 (0.029, 50)

0.095 (0.024, 100)

0.127 (0.047, 500)

0.255 (0.05, 1050)

SCRDA

0.019 (0.036, 651)

0.024 (0.012, 2089)

0.217 (0.041, 587)

0.241 (0.069, 317)

RF

0.109 (0.012, NA)

0.038 (0.009, NA)

0.262 (0.018, NA)

0.21 (0.041, NA)

SQDA

0.005 (0.002, 300)

0.001 (0.001, 300)

0.108 (0.042, 200)

0.001 (0.002, 400)

DLDA2

0.002 (0.001, 600)

0.04 (0.005, 500)

0.224 (0.03, 700)

0.217 (0.055, 600)

DQDA2

0.003 (0.001, 500)

0 (0.001, 600)

0.231 (0.033, 600)

0.224 (0.058, 400)

  1. The reported numbers in each table entry in the form of a (b,c) mean: a is the average prediction error, b is the standard deviation, and c is the median number of predictors selected.