Location: Room 340
Data science in Health Care (Michael Draugelis & Corey Chivers)
This talk will give a brief overview of the Data Science mission at Penn Med and highlight the technical and operational challenges specific to healthcare. Finally there will be a demo of an early view of PatientSignalsML -- a system to accelerate the development and deployment of machine learning solutions for medicine and health care problems. Penn Signals is an integrated platform for processing real-time clinical data streams from a variety of sensors and source systems. The platform provides clinicians and researchers with multiple access-points into this rich stream of data. Penn signals provides the tools needed build, train, test, and, importantly, to deploy predictive applications powered by the data stream. By removing the technical and operational barriers associated with building and deploying predictive applications, Penn Signals aims to accelerate innovation in clinical informatics.
Corey Chivers is the product lead on the data science team at Penn Medicine. He completed his PhD in computational biology at McGill University in 2014 where he specialized in building Bayesian models for predicting large-scale complex dynamic systems. He founded the Montreal R Users group, and blogs about competitive machine learning, probability, and statistics at bayesianbiologist.com.
More details to come
Matt Sundquist (Plotly)
Data Visualization for R, Python, Excel, & MATLAB. Now On The Web.
Plotly is a new web-based platform that lets you use Python, R, MATLAB, and Excel to make beautiful, interactive, web-based graphs in 2D and 3D with D3.js (https://plot.ly/api). In this talk, we'll look at how to use Plotly with IPython, ggplot2, and a web app to make interactive versions of seven of the most famous graphs ever made.
Matt Sundquist previously philosophy at Harvard, wrote for SCOTUSblog.com, worked for the Facebook Privacy Team, and was a Fulbright Scholar in Argentina. He is a Co-founder at Plotly.