Uncertainty-aware INVASE: Enhanced Breast Cancer Diagnosis Feature Selection
This work addresses feature selection for breast cancer diagnosis, offering a domain-specific improvement that is incremental over existing INVASE methods.
The paper tackles the problem of predictive bias in breast cancer diagnosis by introducing an uncertainty-aware INVASE method, which reduces the required queries to about 20% while eliminating almost all bias, compared to nearly 100% queries for uncertainty-agnostic methods.
In this paper, we present an uncertainty-aware INVASE to quantify predictive confidence of healthcare problem. By introducing learnable Gaussian distributions, we lever-age their variances to measure the degree of uncertainty. Based on the vanilla INVASE, two additional modules are proposed, i.e., an uncertainty quantification module in the predictor, and a reward shaping module in the selector. We conduct extensive experiments on UCI-WDBC dataset. Notably, our method eliminates almost all predictive bias with only about 20% queries, while the uncertainty-agnostic counterpart requires nearly 100% queries. The open-source implementation with a detailed tutorial is available at https://github.com/jx-zhong-for-academic-purpose/Uncertainty-aware-INVASE/blob/main/tutorialinvase%2B.ipynb.