Most new predictive modeling packages in R come equipped with cross-validation, but it's also easy to program cross-validation from scratch. There's no excuses for analysts not being attentive to testing and the dangers of overfitting with their models
APR 12, 2011 8:21am ET
To continue, please log in or register.
- Gartners Top 10 Predictions for IT in 2017 and Beyond
- Gartners Top 10 Strategic Technology Trends for 2017
- Virtuous Machines -- How Analytics Will Underpin Artificial Intelligence
- 4 Ways Business Intelligence Software Improves Data Analysis
- Heart Organization Building Big Data Platform to Aid Precision Medicine