The learning algorithm is a process of chaos optimization, which can make the network avoid the local minima problem and false saturation phenomenon. 網(wǎng)絡(luò )的學(xué)習過(guò)程是一種混沌優(yōu)化算法，可有效避免普通神經(jīng)網(wǎng)絡(luò )的局部極值和假飽和現象的發(fā)生。
Based on the congruence non-repetition and ergodicity of chaos, the method will avoid the local optimal solution and find satisfactory globe optimal solution. 由于混沌優(yōu)化算法的疊代具有不重復性和遍歷性，因此該算法可以避免陷入局部最優(yōu)點(diǎn)而獲得全局最優(yōu)。
