Yue Zhao comments

Results 138 comments of


                                            Yue Zhao

Pickle error when using deep learning based models with SUOD

This is a great point! I think the combination comes from two perspectives. First, you may consider using deep learning models as feature extractors, and then apply the classifical OD...

Skip-GANomaly: Skip Connected and Adversarially Trained Encoder-Decoder 这篇文章怎么样呢？

没看过。简单搜了一下，ACCV 2019，有代码 https://github.com/samet-akcay/ganomaly 给Star的人还不少。考虑到代码放出来了，star不少，citation也有20了，感觉风险不大。

Autoencoder learning rate

you make a legit point. I just realize we simply use the default optimizer's learning rate. will relax that...

Autoencoder learning rate

> Hi Lucew, Thanks for the thought on this! Happy to take in a PR to improve the current AE/VAE.

ALAD method implemented

sorry there will be some delay on my end to multiple ddl...will review very soon!! thanks

TypeError: create_target_machine() got an unexpected keyword argument 'jitdebug'

Hi there. sorry for being late. I search around and looks like the solution is to use the following version for specific libraries 1. llvmlite-0.32.1 2. numba-0.49.1

difference between XGBOD and XGBClassifier

The paper is here: https://arxiv.org/abs/1912.00290 To be short, XGBOD is an extended version of xgboost but using unsupervised models to enrich the feature space to achieve a better result. In...

Using Approximate k-Nearest Neighbourhood Queries (Support for Integrating Popular ANN Libraries and Using as a Proximity Backend)

This makes great sense. I used faiss before. I will mark it as a feature to add.

中文版本与英文版本区别不大

中文版内容上应该是完全一致的（最近英文版的更新没有反应到中文版上），如果有兴趣也欢迎交PR让两者同步一下，最近的时间比较紧张 :)

copod od result strange. Especially, continouse od result may be wrong

This is a very good exploration discussion. Any immediate improvement we could add for COPOD to make it more robust? Thanks!