Monday, January 25, 2010

Searching For Ceiling Effects in NASA MDP Components - Part I - Undersampling

Defect distribution visualization can be found here among components. These are verified by components flagged to be used/ignored based on the median number of defects. Components whose defects exceed in number the median are first filtered for selection. These components can be seen below.



Component IDUse/Ignore
22295ignore
22296ignore
22297ignore
22298ignore
22299ignore
22302ignore
22303ignore
22305ignore
22306ignore
24952ignore
24954ignore
24958ignore
24959ignore
24960ignore
24962ignore
24964ignore
24967ignore
24970ignore
24971ignore
25ignore
25492ignore
26ignore
27ignore
28ignore
29ignore
30ignore
31ignore
32ignore
33132ignore
33135ignore
33137ignore
33138ignore
33139ignore
33140ignore
33142ignore
33143ignore
33146ignore
33147ignore
33148ignore
34753ignore
34754ignore
34755ignore
34756ignore
34757ignore
34758ignore
34759ignore
34760ignore
34761ignore
34762ignore
34763ignore
34764ignore
34765ignore
34766ignore
34767ignore
34768ignore
34769ignore
70812ignore
70814ignore
70815ignore
70818ignore
70820ignore
70821ignore
70822ignore
70823ignore
70824ignore
17use
18use
19use
20use
21use
22use
22292use
22293use
22294use
22300use
22301use
22304use
22307use
22308use
22309use
23use
24use
24953use
24955use
24956use
24957use
24961use
24963use
24965use
24966use
24968use
24969use
25493use
33133use
33134use
33136use
33141use
33144use
33145use
33149use
33150use
34752use
70816use
70817use
70819use

Plots of the potential ceiling effects can be found here using subsampling, where M=N (#defective = #non-defective). A 10X10-way cross-validation was used in the experiment.

No comments:

Post a Comment