| Description |
Reflective learning is about algorithms applied to other algorithms to change their make-up and hence, their behavior, in a fashion that improves their performance. It is a form of meta-programming and requires a high level of interest in Artificial Intelligence (AI) and Machine Learning (ML) as a subject of study as well as application.
The purpose of the proposed project is the design, coding and testing of a software platform for reflective learning with immediate applications to image processing, autonomous/genetic/developmental programming, and ML methods used for search and optimization (here, of other programs). During the second phase, there will be interim measurable objectives that would allow the development team to assess its own progress.
The project has three phases: design, iterative coding and testing, and finally, commissioning and documentation. The first phase will be done in close collaboration with the supervising professor; the second relies on precise implementation and meticulous debugging; the final phase prepares for final delivery (including demonstration) of the finished product and the data that proves its proper functioning.
|