tyler-tomita / RandomerForest

Discriminant Projection Forest results, datasets, etc.
44 stars 21 forks source link

1-sided mann-whitney test comparing RF vs. RerF(s+d+r) on benchmark data #52

Closed jovo closed 9 years ago

jovo commented 9 years ago

http://en.wikipedia.org/wiki/Mann%E2%80%93Whitney_U_test

tyler-tomita commented 9 years ago

Wouldn't the signed rank test be more powerful since the Lhats are paired?

jovo commented 9 years ago

good call!

On Fri, May 15, 2015 at 8:51 PM, ttomita notifications@github.com wrote:

Wouldn't the signed rank test be more powerful since the Lhats are paired?

— Reply to this email directly or view it on GitHub https://github.com/ttomita/DPForest/issues/52#issuecomment-102555131.

the glass is all full: half water, half air. openconnecto.me, jovo.me, office hours https://www.google.com/calendar/embed?src=e2ktu4lrgul8anp8hclrcminp8%40group.calendar.google.com&ctz=America/New_York

tyler-tomita commented 9 years ago

p = 0.003. We can just mention this in the text?

jovo commented 9 years ago

yes! and/or put in title of that panel!

On Friday, May 15, 2015, ttomita notifications@github.com wrote:

p = 0.003. We can just mention this in the text?

— Reply to this email directly or view it on GitHub https://github.com/ttomita/DPForest/issues/52#issuecomment-102566853.

the glass is all full: half water, half air. openconnecto.me, jovo.me, office hours https://www.google.com/calendar/embed?src=e2ktu4lrgul8anp8hclrcminp8%40group.calendar.google.com&ctz=America/New_York

jovo commented 9 years ago

how did we do with the corrected RerF

On Thursday, May 28, 2015, ttomita notifications@github.com wrote:

Closed #52 https://github.com/ttomita/DPForest/issues/52.

— Reply to this email directly or view it on GitHub https://github.com/ttomita/DPForest/issues/52#event-316407956.

the glass is all full: half water, half air. openconnecto.me, jovo.me, office hours https://www.google.com/calendar/embed?src=e2ktu4lrgul8anp8hclrcminp8%40group.calendar.google.com&ctz=America/New_York

tyler-tomita commented 9 years ago

Lhat_RF > Lhat_RerF: p = 0.99 Lhat_RF < Lhat_RerF: p = 0.03

We did worse :(

jovo commented 9 years ago

ok. lets repeat with the faster variant and see how we do?

On Thursday, May 28, 2015, ttomita notifications@github.com wrote:

Lhat_RF > Lhat_RerF: p = 0.99 Lhat_RF < Lhat_RerF: p = 0.03

We did worse :(

— Reply to this email directly or view it on GitHub https://github.com/ttomita/DPForest/issues/52#issuecomment-106575655.

the glass is all full: half water, half air. openconnecto.me, jovo.me, office hours https://www.google.com/calendar/embed?src=e2ktu4lrgul8anp8hclrcminp8%40group.calendar.google.com&ctz=America/New_York