PandasUDFs - Scaling Ensembles for Improved Predictions
Offered By: Databricks via YouTube
Course Description
Overview
Discover how to leverage PandasUDFs as a powerful technique for scaling ensemble models in this 38-minute Databricks talk. Learn to transform development code into scalable solutions for category-specific predictions, dramatically reducing runtime from hours to minutes. Explore the general usage, types of PandasUDFs, strategies for overcoming data limits, and equivalent approaches in R and Koalas. Gain insights into applying this method to scale from single models to entire ensembles, enhancing prediction accuracy across diverse categories.
Syllabus
Introduction
The Problem
PandasUDFs
Use Cases
Data Limits
Other Frameworks
Taught by
Databricks
Related Courses
Big Data Analytics in HealthcareGeorgia Institute of Technology via Udacity Model Building and Validation
AT&T via Udacity Maths for Humans: Linear, Quadratic & Inverse Relations
University of New South Wales via FutureLearn Regression Modeling in Practice
Wesleyan University via Coursera Data Science at Scale - Capstone Project
University of Washington via Coursera