Objective Function Analysis models knowledge as a multi-dimensional probability density function (MD-PDF) of the perceptions and responses (which are themselves perceptions) of an entity and an objective function (OF). The learning algorithm is the action of choosing a response, given the perceptions, which maximizes the objective function. The MD-PDF is initially seeded by a uniform random number generator. The response is used to evaluate the OF and the OF is either reinforced or diminished in the probability subspace formed by the perceptions and responses. Stated another way, in response to its perceptions, an entity evaluates every possible value of the OF in the subspace defined by the n perceptions and m responses and chooses the response which maximizes the OF. Eventually the randomly seeded MD-PDF changes to reflect the mathematical properties embodied in the OF.
OFA is demonstrated through several simulations maintained in this repository. Contact phgphd@yahoo.com.
Features
- C++, Qt5 graphics, QwtPlot3D, OpenGL
- Scientific paper, ofapaper.odt (docx)