Questions and answers about philosophy of science, causation, and ...

• cours magistral
• exposé - matière potentielle : multiple causes
• revision

Questions and answers about philosophy of science, causation, and human/machine learning Yuan Qi MIT Media Lab Cambridge, MA, 02139 1 Question (A) Explain how the tools of statistical learning (Bayesian or otherwise) might be useful in thinking about paradigm shifts in science. Is a paradigm shift just the replacement of one model or hypothesis by another, or does it involve something more? (B) What would be the analogy of a Kuhnian paradigm shift in human learning or machine learning? Give at least one specific example of a phenomenon of human learning, and also a phenomenon from machine learning, that can be thought of

• bayesian point of view
• structural causal model
• human learning
• causal support
• bayesian methods
• graphical models
• statistical inference
• hypothesis
• data
• w.w.norton
• w.w. norton
• w. w. norton