A conversation with Quants, Thinkers and Innovators all challenged to innovate in turbulent times!

Join QuantUniversity for a complimentary spring speaker series where you will hear from Quants, innovators, startups and Fintech experts on various topics in Quant Investing, Machine Learning, Optimization, Fintech, AI etc.

Making your Models Explainable

With Oege Dijk from ThoughtWorks

Nowadays it is no longer enough just to worry about getting your models to be performant. If they are going to be deployed and affect real people they also need to be trustable. One major way to increase trust is by making the models transparent and explainable. We are going to look at how you can explain your models and make them understandable to non-coding stakeholders by launching interactive XAI dashboards.

The library includes:

Shap values (i.e. what is the contributions of each feature to each individual prediction?)
Permutation importances (how much does the model metric deteriorate when you shuffle a feature?)
Partial dependence plots (how does the model prediction change when you vary a single feature?
Shap interaction values (decompose the shap value into a direct effect an interaction effects)
For Random Forests and xgboost models: visualisation of individual decision trees
Plus for classifiers: precision plots, confusion matrix, ROC AUC plot, PR AUC plot, etc
For regression models: goodness-of-fit plots, residual plots, etc.

Try the library here: https://github.com/oegedijk/explainerdashboard