Library implementing state-of-the-art Concept-based and Disentanglement Learning methods for Explainable AI

Model explanation provides the ability to interpret the effect of the predictors on the composition of an individual score.

The Codebase for Causal Proxy Model

Python implementation of the goldeneye algorithm to investigate how classifiers utilise the structure of a dataset.

主要包含ModelHelper和NLPHelper，其中ModelHelper主要有特征选择、超参数搜索、模型解释和模型融合等，NLPHelper则是进一步封装了NLP一些常用的操作，常用的网络结构以及几个NLP的任务

SHAP explainer for LightGBM models - Generate feature importance plots, dependence plots, and prediction explanations with one line of code. Make your gradient boosting models interpretable for stakeholders.

A proof-of-concept on how to install and use Torchserve in various mode