📢 انجمن علمی اقتصاد و علوم کامپیوتر دانشگاه تهران برگزار می‌کنند:

🔷 Optimal Learning for Structured Bandits

This study addresses the limitations of traditional multi-armed bandit algorithms by introducing a new algorithm, DUSA, which effectively utilizes structural information to minimize regret. DUSA aligns with the information-theoretic lower bound and is computationally viable, offering minimal regret for both known and new structured bandits.

👤 Negin Golrezaei
Associate Professor, MIT Sloan School of Management

📅 سه‌شنبه ۶ آذر
🕑 ساعت ۱۸:۳۰

🔗 لینک حضور در وبینار
https://meet.google.com/coe-ytav-mzi

@anjomaneghtesad
@CS_Talks_UT