What Skills Do Predictive Modelers Need?
summary
27 years of practice crystallizes into a few concrete claims: a neural‑network variant is used for real‑time missile guidance, automation sifts enormous combinations of features, and SmarterHQ runs tens of thousands of models automatically after heavy engineering work. The practical stack splits between GUI tools and code (R/Python), classic algorithms dominate (regression, trees, k‑NN, ensembles) and one core rule holds: data preparation is more important than the killer algorithm while feature engineering and model ensembling win contests.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 3:57Predictive analytics defined
- 16:12Stats vs predictive modeling
- 27:26Math needed for modeling
- 39:20Skills beyond algorithms
- 44:15Q&A
related talks
56mPredictive Analytics for Practitioners
07mPredictive Modeling and Machine Learning
46mThe AI dominated world – skills required for successfully navigating it
57mGood Practices for Applied Machine Learning
39mForecasting With Classical and Machine Learning Methods
37mLife of a data scientist
32mThe Dangers of Mindless Forecasting
38mTo explain or to predict?