Unlocking the value of your data
summary
Claims that only 19 percent understand the data required for AI and that 1 in 20 models reach production set the stakes, then a Wimbledon use case with IBM Watson turns hundreds of hours into 15‑second highlights but initially overprioritizes top‑ranked players. The pipeline uses OpenScale plus AI Explainability 360 and Fairness 360 to produce contrastive explanations, detect and correct bias, surface under‑seen players, and detect model drift from changing applicant demographics so models are retrained.
AI-written summary · about
chapters (4)
related talks
09mUnlocking the value of your data
12mTrust and transparency of AI for the enterprise
39mBig Data Information Architecture for AI
21mAlgorithms at Scale: Raising Awareness on Latent Inequities in Our Data
16mOperationalizing AI in Finance: Real-Time Analytics
11mAutomation of AI: Accelerating the AI Revolution
26mWhy Machine Learning Algorithms Fall Short...
23mDelivering business success with AI