Towards Global and Human Centered Explanations for Machine Learning Models

In this 26-minute talk from LeadDev, explore the field of Explainable AI (XAI) with Carla Vieira as she presents existing literature and contributions in the field while discussing what remains to be accomplished. Learn about the growing need for interpretable AI systems as these technologies become more prevalent in daily life. Understand the limitations of current AI systems, particularly their lack of transparency and interpretability, which leaves users with minimal understanding of how decisions are made. Discover how XAI systems aim to self-explain their reasoning behind decisions and predictions, and how researchers from various disciplines collaborate to define, design, and evaluate these explainable systems. Gain insights into the importance of global explanations for establishing trust in AI models, despite most current research focusing on explaining individual predictions. This talk serves as valuable reference material for newcomers to XAI, potentially stimulating future research advances while encouraging professionals from other fields to embrace AI benefits without concerns about interpretability.