IEEE ICHI 2019 tutorials: http://www.ieee-ichi.org/tutorial.html
Date/Time: 13:30 – 17:00 on Tuesday June.11.2019
This GitHub repo is set up for the IEEE ICHI 2019 Translational Health Informatics - from Risk Prediction Modeling.
The code/data/notebooks in this repo is specifically used in the Translational Health Informatics Hands-on Tutorial Session.
To help the audience better understand the tutorial, we set up this GitHub repo to host Python code and Jupyter Notebook files for the hands-on session. The audience can read and run the code interactively to know how does the risk prediction process works.
To run the code, users shall
IBM Watson Studio is an integrated environment for data scientists, developers and domain experts to collaboratively work with data to build, train and deploy models at scale. If you are new to Watson Studio, you can find some introduction here.
Notice: It may take several minutes to register and IBMid and get familiar with Watson Studio. If you have time, please try Watson Studio in advance.
Plan B: Binder
We also offer a Plan B which is also easy to use: mybinder.org. By clicking the link above, you will be directed to a webpage which hosts a live JupyterLab service. You can then use the notebooks in that environment (it may take a few minutes for mybinder to prepare the environment).
Notice: If you choose to use binder, all your changes in the browser is temporary and will be gone if you close your browser.
In case you want to try the code on your local computer, please checkout the code to you laptop and setup environment by yourself. It's prefered to have latest version of Python and Jupyter Notebook installed on your laptop.
This tutorial will focus on the issues when utilizing new scientific discoveries in healthcare, especially in the era of Artificial Intelligence, moving from bench to bedside, on to communities and finally to the population. Instructors will firstly give an overview of translational health informatics from risk prediction modeling to risk assessment service, and introduce the pipeline of risk prediction modelling with hands-on, then deep dive to key steps along the translational journey, for addressing real-life problems such as risk model adaption and localization, dealing with unobservable features when applying risk models, and bridging the gap from a risk score to intervention plan.
Contributers of this tutorial are from IBM Research China:
- Jing Mei (organizer, speaker)
- Enliang Xu (organizer, speaker)
- Bibo Hao (organizer, speaker)
- Shaochun Li (organizer, speaker)
- Yuan Zhang (speaker)
- Yiqin Yu (speaker)