we need to defend the statement that "most data mining in analytics does not do tuning"

[timm]

we need to defend the statement that "most data mining in analytics does not do tuning"

can you

• [x] go to https://goo.gl/LjvOkh
• [x] write down the citation counts of the top 50 papers
• [x] sort by those counts
• [x] discard the ones without pdf
• [x] discard the non SE papers
• [x] keep the top 30 survivors

for the surviving 30 papers, fire up in pdf. hunt for tuning, parameter, setting (anything else?) that might suggest they do tunings. also check validity section for any nervousness that their settings are not the wright ones?

paperId	year	citations	mentions tuning at all?	does tuning?
1	1984	310	n	n

we need to defend the statement that "most data mining in analytics does not do tuning"

[timm]

k

[WeiFoo]

Sorted In GOOGLE SHEET

Done! ==>LINK

[timm]

is that sheet public? please make it so.

also, plz add a second worksheet called "how we did it" as a title on the tab and text:

In google scholar, search for "data mining" "software engineering" "defect prediction" Sort by citation Discard the ones without a pdf Discard the non SE papers Keep the top 50 survivors Search for tuning or parameter

[timm]

btw- great work on that sheet

[WeiFoo]

Done!