Meaningful insights: explainability techniques for black-box models on tabular data

Artificial Intelligence (AI) and Machine Learning (ML) are novel data analysis techniques providing very accurate prediction results. They are widely adopted in a variety of industries to improve efficiency and decision-making, but they are also being used to develop intelligent systems. Their success grounds upon complex mathematical models, whose decisions and rationale are usually difficult to comprehend for human users to the point of being dubbed as black-boxes. This is particularly relevant in sensitive and highly regulated domains. To mitigate and possibly solve this issue, the Explainable AI (XAI) field became prominent in recent years. XAI consists of models and techniques to enable understanding of the intricated patterns discovered by black-box models. In this thesis, we consider model-agnostic XAI techniques, which can be applied to Tabular data, with a particular focus on the Credit Scoring domain. Special attention is dedicated to the LIME framework, for which we propose several modifications to the vanilla algorithm, in particular: a pair of complementary Stability Indices that accurately measure LIME stability, and the OptiLIME policy which helps the practitioner finding the proper balance among explanations' stability and reliability. We subsequently put forward GLEAMS a model-agnostic surrogate interpretable model which requires to be trained only once, while providing both Local and Global explanations of the black-box model. GLEAMS produces feature attributions and what-if scenarios, from both dataset and model perspective. Eventually, we argue that synthetic data are an emerging trend in AI, being more and more used to train complex models instead of original data. To be able to explain the outcomes of such models, we must guarantee that synthetic data are reliable enough to be able to translate their explanations to real-world individuals. To this end we propose DAISYnt, a suite of tests to measure synthetic tabular data quality and privacy.

Institutional Processes and Discursive Strategies: Rhetoric and Vocabulary Analysis of CSR and Sustainability

The candidate tackled an important issue in contemporary management: the role of CSR and Sustainability. The research proposal focused on a longitudinal and inductive research, directed to specify the evolution of CSR and contribute to the new institutional theory, in particular institutional work framework, and to the relation between institutions and discourse analysis. The documental analysis covers all the evolution of CSR, focusing also on a number of important networks and associations. Some of the methodologies employed in the thesis have been employed as a consequence of data analysis, in a truly inductive research process. The thesis is composed by two section. The first section mainly describes the research process and the analyses results. The candidates employed several research methods: a longitudinal content analysis of documents, a vocabulary research with statistical metrics as cluster analysis and factor analysis, a rhetorical analysis of justifications. The second section puts in relation the analysis results with theoretical frameworks and contributions. The candidate confronted with several frameworks: Actor-Network-Theory, Institutional work and Boundary Work, Institutional Logic. Chapters are focused on different issues: a historical reconstruction of CSR; a reflection about symbolic adoption of recurrent labels; two case studies of Italian networks, in order to confront institutional and boundary works; a theoretical model of institutional change based on contradiction and institutional complexity; the application of the model to CSR and Sustainability, proposing Sustainability as a possible institutional logic.