
Super Data Science: ML & AI Podcast with Jon Krohn 433: Data Science Trends for 2021
Jan 7, 2021
Ben Taylor, a seasoned data scientist at DataRobot, joins to discuss the pivotal data science trends for 2021 in the wake of COVID-19. He shares his passion for AI and highlights the importance of delivering tangible results through data. The conversation dives into the role of AutoML, the significance of transparent storytelling, and federated learning for preserving data privacy. Ben also addresses AI ethics challenges, the impact of remote work on collaboration, and the rise of emerging software tools revolutionizing the field.
AI Snips
Chapters
Books
Transcript
Episode notes
Crawl, Walk, Run
- New data science teams should prioritize quick wins ("crawl") before tackling long-term projects ("run").
- Allocate the majority of time to projects with clear business value, while reserving some time for innovative explorations.
DataRobot's Focus
- DataRobot, an applied AI leader, offers AutoML and end-to-end ML pipelines.
- They focus on productionizing models, addressing challenges like feature drift and model monitoring.
Transparent Storytelling
- Tools like SHAP, Gradcam, and topic discovery offer insights into black-box models.
- Involving subject matter experts from the outset is critical for interpreting results and identifying potential biases.






