Explainable AI (XAI): Core Ideas, Techniques, and Solutions

Abstract

As our dependence on intelligent machines continues to grow, so does the demand for more transparent and interpretable models. In addition, the ability to explain the model generally is now the gold standard for building trust and deployment of Artificial Intelligence (AI) systems in critical domains. Explainable Artificial Intelligence (XAI) aims to provide a suite of machine learning (ML) techniques that enable human users to understand, appropriately trust, and produce more explainable models. Selecting an appropriate approach for building an XAI-enabled application requires a clear understanding of the core ideas within XAI and the associated programming frameworks. We survey state-of-the-art programming techniques for XAI and present the different phases of XAI in a typical ML development process. We classify the various XAI approaches and using this taxonomy, discuss the key differences among the existing XAI techniques. Furthermore, concrete examples are used to describe these techniques that are mapped to programming frameworks and software toolkits. It is the intention that this survey will help stakeholders in selecting the appropriate approaches, programming frameworks, and software toolkits by comparing them through the lens of the presented taxonomy.

Exploring the Black Box: Analyzing Explainable AI Challenges and Best Practices Through Stack Exchange Discussions

“…Previous studies (Dwivedi et al, 2023;Ding et al, 2022) have identified various XAI tools, and we have also listed 35 tools in Table 1. Among these tools, Shap, LIME, and Eli5 have been consistently highlighted as important tools.…”

“…Additionally, XAI techniques provide versatile methods for explaining complex models; Ribeiro et al (Ribeiro et al, 2016) introduce Lime, an adaptable and expandable approach to explaining predictions in a comprehensible manner, while Lundberg et al (Lundberg and Lee, 2017a) introduce Shap, a powerful framework for interpreting predictions through feature importance values. In the case of available tools in XAI, Previous studies (Dwivedi et al, 2023;Ding et al, 2022) have identified various XAI tools, Among these tools, Shap, Lime, and Eli5 have been consistently highlighted as essential tools.…”

The challenges of integrating explainable artificial intelligence into GeoAI

“…Then there is the when and where one applies XAI: ante hoc XAI (attribution to inputs/outputs in the pre‐modeling phase) and post hoc XAI (post‐modeling phase, e.g., LIME). Researchers also distinguish between white‐ and black‐box methods; the former concerns AI models that are simple and self‐explanatory and the latter considers opaque models and data handling prevalent in DNN (Dwivedi et al, 2023). All are used in GeoAI, as will be seen below.…”

“…Researchers also distinguish between white-and black-box methods; the former concerns AI models that are simple and self-explanatory and the latter considers opaque models and data handling prevalent in DNN (Dwivedi et al, 2023). All are used in GeoAI, as will be seen below.…”

Genetic Programming for Interpretable and Explainable Machine Learning

“…This overview is not meant to be exhaustive, but rather to highlight the various areas where XAI techniques can be applied. For a more intensive survey of current methods in XAI, we direct the reader to recent reviews [11,42] on the topic.…”