小弟撰寫的類神經網路backpropagataion,可以train如xor等互斥問題,使用bcb所完成,因為開發介面較為便利,

大部分使用類別的方法撰寫,所以若有興趣移植到vc的朋友,應該也不會有太大的障礙。-Boy written by neural network backpropagataion, can train such as xor, such as mutually exclusive, the use of BCB completed, because the development of the interface is more convenient, most types of methods used to write, so if interested in transplanted into vc friends, should also will not have too big an obstacle.