Exploring Explainability in Machine Learning

Hey all,

Thanks for all your presentations.

Next week, we will have a retrospective discussion on Friday, 31st January starting at 3:00 PM. Here's the plan for the session:

1. Short Recap of Papers:
   Each one will give a brief recap (less than 5 minutes) of their presentations. 

2. Discussion on Explanation in Machine Learning:

   • Please prepare at least 5 questions or topics related to explanation in machine learning for discussion.
   • We will go around the group to discuss these questions in turns.

3. Report Section:

   • After the discussion, we will talk about the report section and finalize the details.

This session is an excellent opportunity to share insights and clarify any doubts, so please come prepared.

Thank you for your valuable contributions to the session, and I look forward to a productive meeting!

Hanwei

Dear all,

After careful consideration regarding the lunch issue and due to a scheduling conflict with the room, we have made some adjustments to the session plan for Wednesday.

The session will still take place between 10:00 AM and 2:00 PM on E1 3 Room 528, but we will now include a lunch break from 12:00 PM to 1:00 PM. This will allow everyone to relax and refuel before resuming the activities.

Thank you for your understanding and flexibility. Please let me know if you have any questions or concerns. Looking forward to seeing you all there!

Best regards,

Hanwei

Dear all,

We are excited to announce the schedule for the upcoming presentations and retrospective sessions. Please take note of the following dates and times:

Presentation Session 2
📅 Date: Wednesday, January 15
🕙 Time: 10:00 AM – 2:00 PM
📋 Details: Includes 3 presentations

Tulika Nayak – SHAP: Feature Importance Using Shapley Values from Game Theory

• Presentation Chair: Shashank Priyadarshi

Nhi Pham – Counterfactual Explanations: “What-if” Scenarios for Actionable Insights

• Presentation Chair: Mayur Deshmukh

Mayur Deshmukh – Anchors: Rule-based, High-Precision Explanations for Stability

• Presentation Chair: Shreyansh Tripathi

Presentation Session 3
📅 Date: Wednesday, January 22
🕙 Time: 10:00 AM – 12:00 PM
📋 Details: Includes 2 presentations

Shreyansh Tripathi – Graph Neural Networks Explainability: Explaining GNNs for Structured Data

• Presentation Chair: Maurice Vincon

Maurice Vincon – Concept Activation Vectors (TCAV): High-Level, Human-Friendly Explanations

• Presentation Chair: Tulika Nayak

Retrospective Session
📅 Date: Friday, January 31
🕒 Time: 3:00 PM – 5:00 PM
📋 Details: Dedicated time for reflections and feedback

Please ensure to mark your calendars and prepare accordingly. I am looking forward to your active participation and engagement during these sessions.

Best regards,

Hanwei

Hey all,

I am excited to inform you that our first and second presentations will be held as follows:

🗓 Date: January 8th
⏰ Time: 4:00 PM - 6:00 PM
📍 Location: Room E1 3, Room 528

Presentations:

1. Tulika Nayak – SHAP: Feature Importance Using Shapley Values from Game Theory

   • Presentation Chair: Shashank Priyadarshi

2. Shashank Priyadarshi – LRP: Layer-wise Relevance Propagation

   • Presentation Chair: Nhi Pham

Please mark your calendars and join us for these important sessions. We look forward to your participation and engagement!

Best,

Hanwei

Dear all,

We regret to inform you that our first presenter, Juan José Valenzuela, has decided to withdraw from the seminar. As a result, Tulika Nayak and Shashank Priyadarshi will now serve as the first and second presenters. The new order of presentations will follow as listed:

1. Tulika Nayak – SHAP: Feature Importance Using Shapley Values from Game Theory
2. Shashank Priyadarshi – LRP: Layer-wise Relevance Propagation
3. Nhi Pham – Counterfactual Explanations: “What-if” Scenarios for Actionable Insights
4. Mayur Deshmukh – Anchors: Rule-based, High-Precision Explanations for Stability
5. Shreyansh Tripathi – Graph Neural Networks Explainability: Explaining GNNs for Structured Data
6. Maurice Vincon – Concept Activation Vectors (TCAV): High-Level, Human-Friendly Explanations

Currently, we lack sufficient input to finalize the time slots for the first two presentations. Therefore, we kindly request that everyone indicate their availability for the following dates:

• December 16–20: https://doodle.com/meeting/participate/id/dGzZ74rb
• January 6–10: https://doodle.com/meeting/participate/id/en7ZVvPa

Please submit your availability by tomorrow so we can finalize the schedule for the first and second presentations. Your prompt response is greatly appreciated to ensure the seminar runs smoothly.

Thank you for your cooperation.

Best regards,

Hanwei