Data Analytics: Realizing the Full Potential
summary
Central claim: process beats hype — start with a concrete problem and prototype interfaces before optimizing models. Examples: Google query logs deliver a two-week head start on flu detection; a Kaggle Heritage Health prize leaderboard chases tiny metric gains and fails to meet the stated goal; an e-discovery system for Dages uses predictive coding and accompanying lawyer-facing accuracy tools to enable over 99% reduction in manual review; design-led work at Procter & Gamble yields the Swiffer from user observation.
AI-written summary · about
chapters (9)
- 0:00Introduction
- 3:13Big data hype vs reality
- 7:46Data science process limits
- 11:54Kaggle and tool obsession
- 18:11Start with the problem
- 25:57E-discovery case study
- 37:00P&G adoption case study
- 42:19Human-centered data design
- 46:35Call to action
related talks
28mGenerate Insights, Not Excuses: How your data is slowing you down
01h 01mYou don't need AI; You need to know what your data is.
24mValue-Driven Analytics Development
38mDo's and don'ts of big data analytics
41mProficiency + Creativity = Happy Data Practitioners
27mNavigating Data's Next Great Shift
26mData & Analytics Strategy: 5 Overlooked Components
23mBig data and big insights