Consistent Model Selection by an Automatic "Gets" Approach.
<p>We establish the consistency of the selection procedures embodied in <em>PcGets</em>, and compare their performance with other model selection criteria in linear regressions. The significance levels embedded in the <em>PcGets</em> Liberal and Conservative algorithms...
| Main Authors: | , , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
Blackwell Publishing Ltd
2003
|
| Summary: | <p>We establish the consistency of the selection procedures embodied in <em>PcGets</em>, and compare their performance with other model selection criteria in linear regressions. The significance levels embedded in the <em>PcGets</em> Liberal and Conservative algorithms coincide in very large samples with those implicit in the Hannan-Quinn (HQ) and Schwarz information criteria (SIC) respectively. Thus, both <em>PcGets</em> rules are consistent under the same conditions as HQ and SIC. However, <em>PcGets</em> has rather different finite-sample behaviour. Pre-selecting to remove many of the candidate variables is confirmed as enhancing the performance of SIC.</p> |
|---|