Speaker:  Yutian Chen, Staff Research Scientist, Deepmind

Time: 10:00-11:30 a.m., August 15, 2023, GMT+8

Venue: Tecent Meeting ID: 563 818 373

Abstract:

Meta-learning for optimization, also known as learning to optimize, aims to develop enhanced optimization algorithms that surpass the efficiency of conventional optimization methods for specific machine learning tasks. In this presentation, I will delve into the core principles of meta-learning and its applicability to zero-order optimization scenarios. We will also address the primary challenges encountered in this field and showcase solutions from our latest research. These span the spectrum from adopting parameterized versions of standard evolutionary algorithm to crafting neural black-box optimizers from scratch.

Biography:

Dr Yutian Chen is a staff research scientist at DeepMind. He obtained the B.E. of Electronic Engineering from Tsinghua University, and PhD in machine learning at the University of California, Irvine, and later worked at the University of Cambridge as a research associate (Postdoc) before joining DeepMind. Yutian took part in the AlphaGo and AlphaGo Zero project, developed Game Go AI programs that defeated the world champions. The AlphaGo project was ranked in the top 10 discoveries of the decade 2010s by the New Scientist magazine. Yutian has conducted research in multiple machine learning areas including Bayesian methods, offline reinforcement learning, generative models and meta-learning with applications in gaming AI, computer vision and text-to-speech. Yutian also serves as area chairs for multiple academic conferences including AISTATS, ICLR, NeurIPS and ICML.

Source: PKU-IAI