期刊名称：International Journal of Wireless & Mobile Networks
印刷版ISSN：0975-4679
电子版ISSN：0975-3834
出版年度：2020
卷号：12
期号：3
页码：31-44
DOI：10.5121/ijwmn.2020.12303
出版社：Academy & Industry Research Collaboration Center (AIRCC)
摘要：As the telecommunication network components and functions are getting commoditized, the complexity in configuration and optimization increases. Several automation techniques are evolving from traditional deterministic algorithms (pre-defined rulesets obtained from experience accumulated by humans) that were heuristic-based to more cognitive and stochastic-based algorithms. The aim of this paper is to introduce the seven layers in wireless telecommunication networks that uses stochastic or AI algorithms, explain the need for monitoring and possible potential biases in each layer of the stochastic algorithm stack and finally conclude with evaluation methods, techniques for detecting false positive and false negative proposals in autonomous network functions. The main subject of the paper is to provide a background on the need of explainable AI for autonomous network functions. The paper includes introduction of two models ANOBIA and INFEROBIA models that helps to achieve explainable AI for autonomous network functions in wireless and mobile networks.
关键词：Explainable AI;Machine Learning;Artificial Intelligence;Precision;Recall;BIAS;Variance;Algorithm and Mitigation methods
