论文部分内容阅读
混合型学习模型HLM将概念获取算法HMCAP和神经网络算法FTART有机结合，能学习多概念和连续属性，其增量学习算法建立在二叉混合判定树结构和FTART网络的基础上，在给系统增加新的实例时，只需进行一遍增量学习调整原结构，不用重新生成判定树和神经网络，即可提高学习精度，速度快、效率高．本文主要介绍该模型中的增量学习算法．
The hybrid learning model HLM combines the concept acquisition algorithm HMCAP with the neural network algorithm FTART to learn multiple concepts and continuous attributes. The incremental learning algorithm is based on the binary mixed decision tree and FTART network, When a new instance is made, incremental learning is only needed to adjust the original structure. Without having to regenerate the decision tree and the neural network, the learning precision can be improved, the speed is fast and the efficiency is high. This article mainly introduces the incremental learning algorithm in this model.