No, I never published it. Here you go: https://buttignon.next-cloud.org/index.php/s/9x9E7o7MQHGaHnt
If you have any question, let me know.
1 Like
Hi @pizzatakeaway . Thank you for the link. It was realy interesting to read.
I still have a few questions: Did you run your algorithms (uHTM rather than mHTM) on NAB? If yes, how were the results?
I wonder if the changes (e.g. the Confusion Matrix) have any influence.
Since you didn’t publish the work I assume you didn’t publish your code either?
I am currently writing my MA and trying to combine HTM with a prefilter ( analog to contrast enhancement in the eye). In the context of this, I have also dealt with multivariate analysis, but it is not my main focus.
1 Like
Hi! Sorry for the late reply.
What I defined as uHTM is actually the univariate-HTM used in NAB (as well the other univariate algorithms). Since all the data in NAB are univariate I did not run any experiment with them (Subutai, Scott and Alexander already wrote a paper on it).
The code is still waiting for a refactoring 
Can you share your code for multivarient model?