search

benmack / oneClass

One-class classification in the absence of test data.
GNU Affero General Public License v3.0
31 stars 12 forks source link

Reproduce example from Can I trust my one class classification #3

Closed waldnerf closed 5 years ago

waldnerf commented 5 years ago

Hi Benjamin I am facing a one class classification problem and would like to use the approach you developped in your paper entitled "Can I trust my one-class classification?". The current vignette does not seem to explain how to create the diagnostic plots (Fig1). Could you help me with that? Cheers Franz

benmack commented 5 years ago

Hi Franz, thank you for your interest. The full Fig1 from the paper is not implemented in the package like shown there, however, the most important part for me was the histogram plot in combination with the boxplots. The densities lines (which exist in the paper but not in the vignette) and the boxplots/histogram (which exist in the vignette) both represent the distribution of the same data. I should somewhere have the code with which I created Fig1. However, you have to note that especially a.1 and a.2 of Fig1 can only be created in case of a two-dimensional feature space and serve only for understanding purposes in the paper. I can try to find the code for Fig1 at the weekend. Let me know if you think if it might help you and I can try to find it and publish it e.g. as gist.
Best regards Ben