Project

This is a review of my previous work of master capstone project.

Two algorithm

• regression
• smoothed l0 norm

Five methods of power comparison

• top n
• type 1 error
• permutation (5% lagest beta)
• permutation (largest beta)
• permutation (mean square error)

Calculation

• type one error

1. reg

##        [,1]   [,2]   [,3]  [,4]  [,5]
## [1,] 1.0000 1.0000 1.0000 1.000 1.000
## [2,] 0.9700 1.0000 1.0000 1.000 1.000
## [3,] 0.9667 1.0000 0.9967 1.000 1.000
## [4,] 0.9300 0.9925 0.9950 0.995 0.995
## [5,] 0.9220 0.9800 0.9840 0.992 0.992

2. smo

##        [,1] [,2] [,3] [,4] [,5]
## [1,] 0.9100    1    1    1    1
## [2,] 0.8900    1    1    1    1
## [3,] 0.8733    1    1    1    1
## [4,] 0.9275    1    1    1    1
## [5,] 0.9020    1    1    1    1

3. difference

##          [,1]   [,2]     [,3]  [,4]  [,5]
## [1,] -0.09000 0.0000 0.000000 0.000 0.000
## [2,] -0.08000 0.0000 0.000000 0.000 0.000
## [3,] -0.09333 0.0000 0.003333 0.000 0.000
## [4,] -0.00250 0.0075 0.005000 0.005 0.005
## [5,] -0.02000 0.0200 0.016000 0.008 0.008

• permutation (5% lagest beta)

1. reg

##        [,1]   [,2]  [,3]   [,4]   [,5]
## [1,] 0.9960 1.0000 1.000 1.0000 1.0000
## [2,] 0.9800 1.0000 1.000 1.0000 1.0000
## [3,] 0.9680 0.9987 1.000 1.0000 1.0000
## [4,] 0.9390 0.9910 0.993 0.9980 0.9950
## [5,] 0.9248 0.9880 0.980 0.9824 0.9928

2. smo

##        [,1]   [,2] [,3] [,4] [,5]
## [1,] 0.9280 1.0000    1    1    1
## [2,] 0.8940 1.0000    1    1    1
## [3,] 0.8760 1.0000    1    1    1
## [4,] 0.8270 1.0000    1    1    1
## [5,] 0.8264 0.9992    1    1    1

3. difference

##         [,1]     [,2]  [,3]   [,4]   [,5]
## [1,] -0.0680 0.000000 0.000 0.0000 0.0000
## [2,] -0.0860 0.000000 0.000 0.0000 0.0000
## [3,] -0.0920 0.001333 0.000 0.0000 0.0000
## [4,] -0.1120 0.009000 0.007 0.0020 0.0050
## [5,] -0.0984 0.011200 0.020 0.0176 0.0072