R Lightning Talks! Time Series, Classifier Interpretation , Reticulate Python

Speaking:

Jeff Newmiller

A Quick Start for Handling Time-Series Data

We will use the task of importing, summarizing and plotting a couple of time-stamped data sets to illustrate some pitfalls and best practices for working with time values in R.

Bio

Jeff Newmiller has been collecting and analyzing data using computers since 1981, and started using R in 2003. He currently assists investors in evaluating risk in solar photovoltaic power system projects, and develops physics-based and empirical (regression-based) models related to solar photovoltaic power equipment performance for DNV GL, an international engineering consulting firm.

_________________________________________________________________________

Stoney Vintson

a. Quick Overview of Classifier Model Interpretation

• LIME (Local Interpretable Model-Agnostic Explanations)
• https://github.com/thomasp85/lime
• https://github.com/marcotcr/lime
• Shapely
• https://christophm.github.io/interpretable-ml-book/

b. Hands on example in R with a github repo for further learning
LIME and SMOTE for dealing with class imbalance in training examples

Bio

Stoney Vintson analyzes image and IMU data at Ceres Imaging. He has worked with computer vision at Trnio and Gigapan as well as a TA at the first immersive data class at General Assembly in San Francisco.

_________________________________________________________________________

Allan Miller

Reticulate: Python for R Users

How to access python modules from inside R using the Reticulate package, co-written by Hadley Wickham of RStudio and Wes McKinney, author of the python pandas library. We will discuss some examples of how to import python modules into R scripts, call python and pandas module functions, and pass data back and forth between python and R.

Bio

An active member of the Bay Area R community sine 2008, Allan Miller is an organizer of the East Bay R Meetup. He teaches statistics and data science using R at UC Berkeley Extension, is a member of their Data Science Advisory Committee, and is a Data Scientist at a local Solar Energy startup.

6:00 Doors open - socializing & individual questions

6:25 Announcements & Introduction

6:30 Speaker(s)

7:45 Wrap up & socializing

8:00 Out the door!