Where is the 'Science' in Data Science?
summary
Argues that data science is not a standalone science but an enabler and that following the scientific method (domain knowledge, hypothesis, measurement, communication, deployment) matters more than mastering frameworks. Shows Florence Nightingale using the Nightingale rose chart to change hygiene practice, Ignai Simlwise running a control experiment on handwashing but failing to persuade leadership, and Gregor Mendel growing 20,000 plants over a decade. Recommends turning notebooks into production via Jupyter notebook → MLOps deployment.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:05Data science is not a science
- 3:35Science behind predictions
- 5:43Scientific method explained
- 13:11Lessons from Florence Nightingale
- 14:15Failure of Ignaz Semmelweis
- 18:10Gregor Mendel’s missed legacy
- 20:55Key traits of a data scientist
- 22:43Full stack data science at Sapient
- 23:25Q&A on explainability
related talks
42mWhy Big Data Is Not Data Science
58mGrowth & change of careers, organizations & responsibility in data science
47mA Practical(ish) Introduction to Data Science
44mBringing Data to L I F E and Life to Data
38mLifecycle of a Data Science Project
24mData Science: the best path to align business objectives
32mThinking Like a Data Scientist
44mPeople, Technology, Culture: How to make Data Science work