Interest-relative induction

This thesis examines the theoretical foundations of Bayesian statistical inference. A more complete mathematical model of inductive inference is developed which takes into account limitations of the agents involved. Recommendations are made for how the process of designing inference algorithms can be improved.