TIMETABLE-CONSTRUCTION WITH MARKOVIAN NEURAL-NETWORK

A neural network which efficiently and nearly optimally solves combina torial optimization problems was applied to the timetable problem. The network was tested on a real-world timetable problem. The results ind icate that the Markovian neural network is an efficient and flexible t ool for solving real-world combinatorial optimization problems such as construction of a timetable for high schools.