Understanding and Visualizing Data Iteration in Machine Learning

要旨

Successful machine learning (ML) applications require iterations on both modeling and the underlying data. While prior visualization tools for ML primarily focus on modeling, our interviews with 23 ML practitioners reveal that they improve model performance frequently by iterating on their data (e.g., collecting new data, adding labels) rather than their models. We also identify common types of data iterations and associated analysis tasks and challenges. To help attribute data iterations to model performance, we design a collection of interactive visualizations and integrate them into a prototype, Chameleon, that lets users compare data features, training/testing splits, and performance across data versions. We present two case studies where developers apply \system to their own evolving datasets on production ML projects. Our interface helps them verify data collection efforts, find failure cases stretching across data versions, capture data processing changes that impacted performance, and identify opportunities for future data iterations.

キーワード
Data iteration
evolving datasets
machine learning iteration
visual analytics
interactive interfaces
著者
Fred Hohman
Georgia Institute of Technology & Apple Inc., Atlanta, GA, USA
Kanit Wongsuphasawat
Apple Inc., Seattle, WA, USA
Mary Beth Kery
Carnegie Mellon University, Pittsburgh, PA, USA
Kayur Patel
Apple Inc, Seattle, WA, USA
DOI

10.1145/3313831.3376177

論文URL

https://doi.org/10.1145/3313831.3376177

動画

会議: CHI 2020

The ACM CHI Conference on Human Factors in Computing Systems (https://chi2020.acm.org/)

セッション: Machine learning & state detection

Paper session
316A MAUI
5 件の発表
2020-04-28 18:00:00
2020-04-28 19:15:00
日本語まとめ
読み込み中…