Skip to content

Comments

Undo panda vis and merge fix spatial#43

Closed
Zbysekz wants to merge 40 commits intohtm-community:masterfrom
Zbysekz:undo_PandaVis_and_mergeFixSpatial
Closed

Undo panda vis and merge fix spatial#43
Zbysekz wants to merge 40 commits intohtm-community:masterfrom
Zbysekz:undo_PandaVis_and_mergeFixSpatial

Conversation

@Zbysekz
Copy link

@Zbysekz Zbysekz commented Sep 15, 2020

This reverts pandaVis for htm_core nab detector & contains changes in #15 (there was merge conflicts)

breznak and others added 30 commits July 30, 2019 12:47
which is "fake" and non-biological, not HTM related.
Unfortunately, with the current settings the code affected the results
the most.
This reverts commit 09f2839.
now the params are directly comparable to Numenta's.
results with SP w numActiveColumnsPerInhArea
(temporarily) reintroducing spatial anomaly in HTMcore detector
@Zbysekz
Copy link
Author

Zbysekz commented Sep 15, 2020

@breznak this seems good.

I will add thresholds and it should be ready

Copy link
Member

@breznak breznak left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The diffs in HTMcore detector were a bit hard to read, but I tried to check throuroughly and seems good to me 👍 Thanks!

"standard": 63.081419488725054
"reward_low_FN_rate": 76.5626293570994,
"reward_low_FP_rate": 61.359926511549155,
"standard": 71.3094612770284
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

this is with the "master" of htmcore? Are the results stable, ie reasonably consistent over several runs? If so, please also bump our score table in Readme and you can move us to the 1st place! ;)

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

wait a second, i must push the non spatial numbers &try to build with latest htm.core

@breznak
Copy link
Member

breznak commented Sep 15, 2020

@Zbysekz I wanted to keep the conversations and linked PRs working, so I've done merged this PR into the original #15 . So closing here now, and we can continue against #15 and merge that to master when ready.

@breznak breznak closed this Sep 15, 2020
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants