Learning from 8 Years of Data Science Mistakes

Caitlin Hudon shares lessons learned from eight years of data science mistakes in four key areas: 1) technical and analysis mistakes like not evaluating models on held-out data and dropping missing values without consideration; 2) communicating with developers by avoiding jargon and focusing on common goals; 3) communicating with business stakeholders by getting them involved early and framing analyses simply; and 4) infrastructure and team mistakes like lack of documentation and pseudocode. Her advice is to learn communication skills, find your community, and remember everyone is still learning.