A Grey-Box Ensemble Model Exploiting Black-Box Accuracy and White-Box Intrinsic Interpretability

Abstract

Machine learning has emerged as a key factor in many technological and scientific advances and applications. Much research has been devoted to developing high performance machine learning models, which are able to make very accurate predictions and decisions on a wide range of applications. Nevertheless, we still seek to understand and explain how these models work and make decisions. Explainability and interpretability in machine learning is a significant issue, since in most of real-world problems it is considered essential to understand and explain the model's prediction mechanism in order to trust it and make decisions on critical issues. In this study, we developed a Grey-Box model based on semi-supervised methodology utilizing a self-training framework. The main objective of this work is the development of a both interpretable and accurate machine learning model, although this is a complex and challenging task. The proposed model was evaluated on a variety of real world datasets from the crucial application domains of education, finance and medicine. Our results demonstrate the efficiency of the proposed model performing comparable to a Black-Box and considerably outperforming single White-Box models, while at the same time remains as interpretable as a White-Box model.

Discovering Hidden Mental States in Open Multi-Agent Systems by Leveraging Multi-Protocol Regularities with Machine Learning

“…This selection includes some of the major learning paradigms in machine learning. Moreover, this mix of white-box and black-box classifiers will prove in our experimental results that sometimes, white-box models can perform equally well as black-box models, while the former are greatly superior in supporting other extremely significant issues such as interpretability and explainability [ 40 ]. The default parameters are used for these five implementation except the number of layers in the artificial neural network (two instead of one) and the number of classes in SimpleKMeans (three instead of two to cover the three classes in the dataset, i.e., S, F, and N).…”

“…Note that this comparison is possible because the machine learning model employed, a decision tree, is a white-box model. A white-box is a model whose inner logic, workings, and programming steps are transparent, and therefore, its decision-making process is interpretable [ 40 ]. In contrast, a black-box model, such as artificial neural networks, is a model whose inner workings are not known and are hard to interpret [ 40 ], making evaluations such as the one presented in this section impossible.…”

“…As shown in a recent review [ 45 ], most prediction and decision-making methods in intelligent environments are interpretable. Therefore, the use of grey-box models, which combine the interpretability of a white-box model with the accuracy of a black-box model, will be considered [ 40 ].…”

Discovery of Resident Behavior Patterns Using Machine Learning Techniques and IoT Paradigm

“…Figure 7 shows an example of a classification tree that resulted from the data analysis of each of the house devices. The design of the trees was based on those presented by Pintelas et al [46] who developed a semi-supervised methodology grey-box model to explain and understand how the predictive models works achieving a balance between the black box and white box paradigms. The tree shows the discovery of the usage patterns related to turning on and off the lights in the living room using algorithm C4.5.…”

Explainable Machine Learning Framework for Image Classification Problems: Case Study on Glioma Cancer Prediction

“…Developing an accurate and interpretable model at the same time is a very challenging task as typically there is a trade-off between interpretation and accuracy [ 7 ]. High accuracy often requires developing complicated black box models while interpretation requires developing simple and less complicated models, which are often less accurate.…”