You Want the Truth – Adapt Training Domain to Improve Q&A on Technical Text
Uplift Modeling: Making Predictive Models Actionable
The Generalization Paradox of Ensembles
Netflix, Dark Knowledge, and Why Simpler Can Be Better
It is a Mistake to Answer Every Inquiry
Charlie Batch and the Cost of Obfuscation
Be a Data Detective
Haystacks and Needles Anomaly Detection
It is a Mistake to Listen Only to the Data
It Is a Mistake to Extrapolate
It is a Mistake to Discount Pesky Cases
It is a Mistake to Accept Leaks from the Future
Evaluate the Validity of Your Discovery with Target Shuffling
In this paper John Elder describes the power of Target Shuffling to evaluate the validity of your discovery. It’s particularly useful for identifying false positives.
The Ten Most Common Data Science Business Mistakes
It is a Mistake to Rely on One Technique
It is a Mistake to Lack Relevant Data
It is a Mistake to Focus on Training Results
It is a Mistake to Ask the Wrong Questions
Are Orange Cars Really not Lemons