Zhangzhi Xiong (熊章智)

• Deployed classical AI agent algorithms (Minimax with Alpha-Beta Pruning) for the Halma game.
• Implemented Neural Approximate Q-learning (DQN) to enhance the agent's decision-making capabilities.

• Modeled chemical bond information as additional inductive bias for better molecular representation.
• Designed Cross-Attention mechanisms and Multi-head GNNs for effective multi-feature fusion.
• Conducted extensive ablation studies to verify the effectiveness of the novel structural designs.

• Images and pictures are lovely! It's also important and beneficial to develop the passion to introduce and demonstrate our work to others in the community! - A.A.Efros
• Codes are not made public due to course regulations.

• Conducting free explorations on factors that affect Catastrophic Forgetting Phenomenon.
• Design curriculum accordingly to mitigate Catastrophic Forgetting and then attempt on verifying validicity of curriculum transfer.

🎉 Paper is awarded an Oral Presentation at the ECML-PKDD MIDAS Workshop! It's a pity that we can't make it to Naples, Italy, to participate and have to give up our submission.

• Extreme Volatility formulation and Warning via machine learning & media source impact analysis.