LEARNINGCORE
Diversity Prediction Theorem
Crowd error equals average error minus diversity. Diverse models beat the lone genius.
How it works
- Crowd error = average individual error − predictive diversity.
- Diverse almost-right models cancel each other's misses.
- One genius loses to a decent, disagreeing committee.
Use → Ensembles, forecasting teams, model committees.
Watch out → Diversity of faces with uniformity of models — the theorem needs model diversity.
Try it → Average three different forecasts instead of picking your favorite.
REDCAPE jobs → Predict · Design · Communicate
Free sources → Coursera M11 (Prediction & the Many-Model Thinker)
Attach to a lattice — combines with
Different families, shared jobs. Disagreement between them is where insight lives.
DIFFUSION
Bass Diffusion (S-Curve)
Innovators plus imitators make the S-curve. Spend early on seeds, later on social proof.
SHARED JOBS: PREDICT · DESIGN
LINEAR & NONLINEARCategorical & Threshold Models
Buckets and cutoffs turn messy attributes into testable if-then rules.
SHARED JOBS: PREDICT · COMMUNICATE
DECISIONSDecision Trees + Value of Information
Fold back probabilities; pay for information only if it changes the action.
SHARED JOBS: DESIGN · COMMUNICATE
More in Learning
Replicator Dynamics (Fisher)
Winning strategies reproduce faster; adaptation speed rises with variation.
Multi-Armed Bandit
Balance pulling the best-known lever against testing new ones. Regret is the score.
Rugged Landscapes (NK)
Interacting parts create many local peaks; hill-climbers get stuck, diverse searchers find higher ground.