One of the key bottlenecks in building ML systems is creating and managing the massive training datasets that today’s models learn from.
Alex Ratner outlines work on Snorkel, an open source framework for building and managing training datasets, and details three key operators for letting users build and manipulate training datasets: labeling functions for labeling unlabeled data, transformation functions for expressing data augmentation strategies, and slicing functions for partitioning and structuring training datasets. These operators allow domain expert users to specify ML models via noisy operators over training data, leading to applications that can be built in hours or days rather than months or years. Alex explores recent work on modeling the noise and imprecision inherent in these operators and using these approaches to train ML models that solve real-world problems, including a recent state-of-the-art result on the SuperGLUE natural language processing benchmark task.
- A basic understanding of machine learning
What you'll learn
- Discover learning techniques for building, managing, and iterating on training datasets and modeling pipelines for ML in general and using the Snorkel framework
This session is from the 2019 O'Reilly Artificial Intelligence Conference in San Jose, CA.
- Title: Building and managing training datasets for ML with Snorkel
- Release date: February 2020
- Publisher(s): O'Reilly Media, Inc.
- ISBN: 0636920370796
You might also like
40 Algorithms Every Programmer Should Know
Learn algorithms for solving classic computer science problems with this concise guide covering everything from fundamental …
Head First Design Patterns, 2nd Edition
You know you don’t want to reinvent the wheel, so you look to design patterns—the lessons …
Kubernetes for Absolute Beginners - Hands-on
Learn Kubernetes in simple, easy and fun way with hands-on coding exercises. For beginners in DevOps. …
Fundamentals of Software Architecture
Salary surveys worldwide regularly place software architect in the top 10 best jobs, yet no real …