Students should be able to:

• Explore the Motivation and Definition of Explainability
  • Understanding the importance of explainability in AI systems.
  • Defining related terms such as interpretability and transparency.
• Achieve a solid background on accounts of Explanation in Social Sciences Literature
  • Examining the primary perspectives on explanation in social sciences.
  • Drawing insights from existing theories and frameworks.
• Understand local and Global Explanation Methods
  • Differentiating between local and global explanation techniques.
  • Exploring model-agnostic methods for generating explanations.
• Interpret Models and Post-hoc Explanations
  • Investigating the concept and utility of interpretable models.
  • Understanding post-hoc explanations and their role in XAI.
• Explain One-shot Decision and Sequential Decision Making Models
  • Contrasting explainability in one-shot decision scenarios.
  • Analysing the unique challenges of sequential decision making models.
• Explain in Embodied and Non-Embodied AI Systems
  • Investigating the role of explainability in AI systems with physical embodiment.
  • Analysing the implications of explainability in non-embodied AI systems.
• Understand the Role of Causality and Interactivity in XAI
  • Understanding how causality influences explainability in AI systems.
  • Exploring the significance of interactivity for generating explanations.

• Implement of Explainable Machine Learning Techniques
  • Understanding and implementing techniques such as LIME, SHAP, PDP, CNNs with attention, and GANs with explainability.
• Be familiar with XAI Python Libraries
  • Getting acquainted with popular XAI libraries like Captum for practical implementation.
• Evaluate of XAI Methods
  • Applying quantitative and qualitative metrics to evaluate XAI methods.
  • Incorporating human-centric evaluations and conducting user studies.
• Apply XAI Methods to Real-World Datasets
  • Applying XAI techniques to real-world datasets for practical insights and interpretation.

• Effectively work with team members from diverse backgrounds and cultures.
• Demonstrate proficiency in project planning and management.
• Effectively communicate technical ideas to audiences with varying levels of expertise.
• Ensure Reproducibility and Collaborative Work
  • Implement best practices for reproducibility in research and project work.
  • Utilise tools and platforms for collaborative work and sharing code efficiently (i.e., Github).