Cognitive models

All Maia’s work involves activities where people make decisions and perform actions. For this we model behaviour in relation to how real humans behave. In our models we aim to replicate the ability of people to observe their environment, to collect information, to make sense of it, to remember, to reason, to make decisions and to act. Our aim is always to do that for a very specific sphere of human activity using all the power and reliability of computers.

In the Legion pedestrian simulator the entities that represent people were developed after extensive observation and study around the world to understand how real pedestrians behave. This understanding was then integrated into the model that underlies the Legion simulator.

Financial markets are dynamic systems driven by the behaviour of their human and robotic participants. Our work there has been to reverse engineer the market process and so to understand the patterns of behaviour of the participants. Our aim is to learn how the participants perceive the market and react to it, and then to create algorithms with autonomous agents that have the ability to perceive, integrate data and make decisions in similar ways.