This thesis presents the conceptual development and an operational prototype of an innovative modelling framework for the transit assignment problem, structured in a multi-agent way and inspired by a learning-based approach. The proposed framework is based on representing passengers and both their learning and decision-making activities explicitly. The underlying hypothesis is that individual passengers are expected to adjust their behaviour according to their knowledge and experience with the transit system performance. An operational prototype was implemented to model the transit assignment process in the morning peak period. Using Reinforcement Learning to represent passenger's behavioural adaptation and accounting for differences in passenger's preferences and the dynamics of the transit network, the prototype has demonstrated that the proposed approach can simultaneously predict how passengers dynamically choose their routes and home departure time, and estimate the total passenger travel cost in a congested network, as well as loads on different transit routes.