In this release, there are multiple exciting new models are introduced, including:

• MCD
• CBLOF
• AutoEncoder

Several performance optimizations are also implemented:

• numba
• Parallelization for multi-core support in certain models

Besides, pyod is officially supporting Python 3.7 now. Multiple incremental changes are also in this release, and some corresponding updates due to the dependent library changed (sklearn LOF model) are also included.

Last but not least, welcome Zain Nasrullah to become a core developer for pyod. We are preparing a paper for JMLR. Hopefully, we could refer and cite the library shortly.

A new figure for selected models ->