Next Meetup

Deep Learning Study: Neural Architecture Search (NAS) for AutoML
AutoML is an exciting development of AI in 2018. In January this year, Google released AutoML Vision. Then in July Google launched AutoML for machine translation and natural language processing. Both packages have been used by companies for practical applications. We will study the 2 papers on the Neural Architecture Search (NAS and NASNets), invented by Barrent Zoph and Quoc Le at Google Brain. This algorithm uses a recurrent network to generate the model descriptions of neural networks and train this RNN with reinforcement learning to select the best neural network. The whole learning process is automated, without human design. In its newest implementation (NASNets), it achieves better accuracy than the best human-invented architectures for ImageNet dataset while having 28% reduction in computational demand, and surpasses state-of-the-art by 4.0% on the COCO dataset. Paper to read: 1. Barret Zoph et al, "Neural Architecture Search with Reinforcement Learning" ICLR 2017, https://arxiv.org/abs/1611.01578 2. Barret Zoph et al, "Learning Transferable Architectures for Scalable Image Recognition", CVPR 2018, https://arxiv.org/abs/1707.07012 Background reading: 1. Junling Hu, "Understanding AutoML and Neural Architecture Search", AI Frontiers Medium Publication. https://medium.com/aifrontiers/understand-automl-and-neural-architecture-search-4260a0942116 This discussion will be led by Junling Hu Bio: Junling Hu is the founder and CEO of question.ai. She is also the Chair of AI Frontiers Conference (https://aifrontires.com). Before starting her company, she was Director of Data Mining at Samsung, where she led a team to build large-scale recommender systems. Prior to Samsung, Dr. Hu led data science teams at PayPal and eBay, providing machine learning solution to company-wide operation, ranging from product search ranking, sales prediction, user opinion mining to targeted marketing. Dr. Hu has more than 1,000 scholarly citations on her papers. She is a recipient of CAREER award from National Science Foundation, for her work on Multi-agent Reinforcement Learning. She holds a Ph.D. in AI from University of Michigan at Ann Arbor. Agenda: 6:30–7 pm Meet and greet 7-8 pm Presentation and discussion 8-8:30pm Social and catch up

TechCode

1172 Castro St · Mountain View, CA

2 comments

Upcoming Meetups

What we're about

We are a professional organization for AI practitioners in the Silicon Valley. We aim to bring together data scientists, engineers, and business people working in AI and big data area. We host seminars, interactive group meetings, and mentoring sessions. We provide an exchange platform for big data professionals to share their experiences, learn about the newest technologies and explore potential startup opportunities. Join us today. Find like-minded people in AI and grow your career and AI and big data business with us.

Members (5,805)

Photos (72)

Find us also at