SAS システム

The QUANTSELECT Procedure

Model Information
Data Set WORK.BASEBALL
Dependent Variable Salary
Selection Method Adaptive LASSO
Quantile Type Single Level
Stop Criterion AIC
Choose Criterion SBC

Number of Observations Read 322
Number of Observations Used 263

Class Level Information
Class Levels Values
Div 4 AE AW NE NW


Quantile = 0.1



SAS システム

The QUANTSELECT Procedure
 
Quantile = 0.1

Selection Summary
Step Effect
Entered
Effect
Removed
Number
Effects
In
AIC AICC SBC Adjusted
R1
0 Intercept   1 2008.3489 2008.3642 2011.9211 0.0000
1 crRuns   2 1918.7675 1918.8137 1925.9118 0.1566
2 nHits   3 1897.2425 1897.3352 1907.9590* 0.1904
3 yrMajor   4 1897.2476 1897.4027 1911.5362 0.1903
4 crBB   5 1896.1765 1896.4100 1914.0373 0.1919
5 nBB   6 1894.1257 1894.4539 1915.5587 0.1950
6 Div AW   7 1891.0577 1891.4969 1916.0628 0.1996
7 crHome   8 1892.2575 1892.8245 1920.8348 0.1977
8   crHome 7 1891.0577 1891.4969 1916.0628 0.1996
9 nAtBat   8 1889.6365* 1890.2034* 1918.2137 0.2017*
10 nHome   9 1891.2595 1891.9710 1923.4089 0.1991
11 crHome   10 1890.7463 1891.6194 1926.4679 0.1998
12 nRBI   11 1892.7421 1893.7939 1932.0358 0.1966
13 Div NW   12 1892.9745 1894.2225 1935.8404 0.1961
14   nRBI 11 1891.3217 1892.3735 1930.6154 0.1988
15 crAtBat   12 1891.8810 1893.1290 1934.7468 0.1978
* Optimal Value Of Criterion

Selection stopped at a local minimum of the AIC criterion.

The model at step 2 is selected where SBC is 1907.96.

Panel showing how the standardized coefficients and the CHOOSE= criterion change with the effect sequence.

Panel showing how selection criteria change with the selection step.

Plot showing how the average check loss changes with the effect sequence.



SAS システム

The QUANTSELECT Procedure
 
Quantile = 0.1

Selected Effects: Intercept nHits crRuns

Parameter Estimates
Parameter DF Estimate Standardized
Estimate
Intercept 1 -4.397043 0
nHits 1 0.878564 0.085678
crRuns 1 0.327350 0.242977