Explore data preparation techniques for analysis in R through this comprehensive conference talk from ODSC West 2015. Learn how to detect and fix common data quality issues, automate routine steps, and improve the success rate of data science projects. Follow along with interactive demonstrations using R, RStudio, and essential packages as John Mount and Nina Zumel guide you through the fundamentals of data preparation. Discover practical solutions for handling missing values, novel categorical levels, and variable treatment. Gain insights into creating effective treatment plans and improving model quality. Access downloadable materials to practice and reinforce your learning at your own pace.
Overview
Syllabus
Intro
Agenda
Workshop Overview
Data Preparation
Typical Data Problems
Example
We are dead values
Identifying missing values
Alternative approaches
Missing values
Pragmatic solution
Novel categorical levels
Training data
Problem statement
Capital categorical variables
Btree solution
Categorical variables
Using impact
Equity
Indicator Variables
Example Package
Treatment Path
Treatment Plan
Model Quality
Variable Treatment
Bad Days
Taught by
Open Data Science
