A Big-Bang Big-Crunch Type-2 Fuzzy Logic System For Generating Interpretable Models In Workforce Optimization
2018 IEEE INTERNATIONAL CONFERENCE ON FUZZY SYSTEMS (FUZZ-IEEE)(2018)
摘要
Efficient utilization of resources (engineers) is critical to optimal service delivery in service-based industries, such as water, electricity or telecom companies. One of the ways in which efficiency can be improved is by optimizing the geographic area in which the engineers operate. This process is known as Work Area (WA) optimization and it is a sub-domain of workforce optimization.In previous attempts to tackle the work area optimization problem, various machine learning algorithms have been proposed to optimize the work areas. However, they don't provide much insight into when and which WAs must be optimized. This is an important question, as optimizing a WA which is already optimum can lead to a loss in efficiency as the engineers can take time to adjust to the new WA.This paper presents a Type-2 Fuzzy Logic System (FLS) which has been optimized by the Big-Bang Big-Crunch approach to allow maximizing the model interpretability and allow good prediction for the future performance of WAs. We can then use these predictions to determine which WAs must be optimized.We compare the proposed type-2 FLS model against a type-1 counterpart, a stacked autoencoder deep neural network and single hidden layer neural network.The results show that the proposed type-2 FLS provides a highly interpretable model which predicts the future performance of WAs with a reasonable error rate. In addition, it also provides the necessary insight into which parameters of the WA determine the future performance. This allows answering the question of when and which WAs must be optimized.
更多查看译文
关键词
Interval Type-2 fuzzy logic, Big-Bang Big-Crunch, Stacked Autoencoder Deep Learning
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络