Open Access Open Access  Restricted Access Subscription or Fee Access

High-Dimensional Explainable AI for Cancer Detection

Jinying Zou, Feiran Xu, Yuyi Zhang, Ovanes Petrosian, Kirill Krinkin

Abstract



The real industrial problems are often influenced by many uncertain factors, leading to computing issues by AI solutions, such as using medical data to predict cancer. As a new field, explainable AI still faces many challenges, such as the high-dimension of features to ex- plain. This paper proposes two possible solutions to the high-dimension XAI problem. The first solution is the Bi-level approach based on the method from cooperative coalitional
game theory. In order to solve the high-dimension problem, the features are clustered using the historical input data set. In particular, the euclidean metric and size-constrained k-means algorithm are used. The second approach is based on the sampling algorithm for calculating the Shapley value, which makes it possible to compute the approximation of the Shapley value without using the coalitional game theory and clustering approaches. This paper implements the XAI solution in cancer prediction algorithm based on a Sampling approach and Bi-level approach, which is new in XAI field.

Keywords


Explainable AI, Anomaly detection, High-dimension, Bi-level approach, Sizeconstrained clustering, Sampling approach.

Full Text:

PDF


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information. Also: DOI is paid service which provided by a third party. We never mentioned that we go for this for our any journal. However, journal have no objection if author go directly for this paid DOI service.