JVN SEMINAR - 25th December 2018
JVN SEMINAR
...............................................
AI BRINGS DOCTORS TO YOUR HOME
&
BIG DATA & DATA SCIENCE AT MOMO: WHEN DATA SCIENCE MEETS ENGINEERING
- Thời gian: 09:00 - 11:00 am, thứ Ba 25/12/2018
- Địa điểm: Phòng Lecture Hall, Viện JVN
- Nội dung:
1. Huy Phan và Oliver : 09:00 - 09:45am
Title: AI Brings Doctors to Your Home
Abstract: The ongoing demographic change has resulted in a large aging population with increased life expectancy in many countries around the world. This imposes tremendous burdens on the health and medical care systems as well as the socio-economic structure of these countries. Consequently, it raises an urgent need for efficient, accessible, cost-effective, and scalable healthcare solutions that can serve a large number of aging people. While developing such solutions could potentially relief the habitants’ dependence on in-hospital treatment and improve their safety, it also requires preserving their privacy meanwhile maintaining comfort.
In our team at Oxford, we are exploring state-of-the-art artificial intelligence combined with non-invasive and wearable sensors, actuators and modern communication technologies aimed at transitioning healthcare solutions from hospital-oriented to home-based environment. In simpler terms, we are developing user-friendly wearable devices, new signal processing techniques, and novel machine learning algorithms targeted at home-based health and medical care monitoring applications. In this talk, we will showcase three research projects we are currently working on to achieve these goals. They are: (1) acoustic monitoring for ambient assisted living; (2) brain state monitoring with wearable devices for mental health and sleep anomalies; and (3) Parkinson disease diagnosis and assessment using smartphones.
2. Dũng Hồng : 09:45 - 10:30am
Title: Big Data & Data Science at MoMo: When Data Science meets Engineering
Abstract: It takes a lot of work to serve a Machine Learning model on a live production system. The "grunt work" does not just consist of preparing, collecting and cleaning the data. When a model is packed nicely and all offline metrics look great, the real work begins, with the goal of serving the model on production.
At MoMo, we take end-to-end Data Science seriously. Once the model is validated by offline metrics, everything must be A/B tested. It then takes a lot of effort to analyze the results and get insights from the experiments. Eventually, the whole pipeline is pushed to production, where data and models must be processed automatically.
This talk will present various parts of the ML pipelines at MoMo, from the tools that we built, to the processes and best practices that we experimented. We will also talk about our principles in applying Machine Learning & Data Science and how it works for a startup environment like MoMo.
LINK ĐĂNG KÝ: https://goo.gl/forms/QnvTMmuBl2tL8Rr12
SPEAKER SLIDES:
https://drive.google.com/file/d/1zOB5i1spEQUt5r2ktNbnQnfGxPgHhIdc/view?usp=sharing
https://drive.google.com/file/d/1zSuy84zzDvSktByHgcW_csc6g9WOQ3Y6/view?usp=sharing