What is it about?
This paper covers an experiment which demonstrates how the usage of explainable AI methods can help improve the performance of predictive models in telecom networks. Use of SHAP, an explainable AI tool, allows an analyst to improve predictions in near real time.
Featured Image
Photo by Luke Chesser on Unsplash
Why is it important?
This is useful because previously, research has not tried to integrate an explainable "Human In-The-Loop" approach to predictive, near real time modelling. We demonstrate a method of doing so and improving performance.
Perspectives
We notice that SHAP suits a model-agonistic method of achieving human in-the-loop near real time predictive modelling, which is a novel use case. In the future, we may see an expansion of Shapley values or another XAI library to expand the capabilities of time series analysis for our purposes.
James Barrett
University of the West of England
Read the Original
This page is a summary of: Analyst-Driven XAI for Time Series Forecasting: Analytics for Telecoms Maintenance, May 2024, ACM (Association for Computing Machinery),
DOI: 10.1145/3674029.3674035.
You can read the full text:
Contributors
The following have contributed to this page
