Focus on improving judgment rather than predicting precise outcomes. Good judgment means updating beliefs with evidence, not ignoring it or relying on gut feelings.
Bayesian thinking updates beliefs with new evidence using prior probabilities and likelihoods to form posterior probabilities.
Differential privacy adds noise to protect individual data while allowing useful aggregate analysis. ε (epsilon) is the privacy budget - lower means more privacy.