多层感知器学习算法可研报告.docVIP

多层感知器学习算法研究 中文摘要 多层感知器是一种单向传播的多层前馈网络模型，由于具有高度的非线性映射能 力，是目前神经网络研究与应用中最基本的网络模型之一，广泛应用于模式识别、图 像处理、函数逼近、优化计算、最优预测和自适应控制等领域。而多层感知器采用的 是BP算法。BP算法的收敛速度慢是个固有的缺点，因为它是建立在基于只具有局 部搜索能力的梯度法之上的，是只具有局部搜索能力的方法，若用于多个极小点的目 标函数时，是无法避免陷入局部极小和速度慢的缺点的。因此，对BP算法的研究一 直以来都是非常重要的课题。 毕业设计课题旨在对多层感知器的学习算法进行研究，并提出一种新的学习算 法。由于BPWE (权值外推BP)算法和TBP (三项BP)算法都是基于权值调整的改 进算法，而考虑将TBP算法中的均衡因子融入到BPWE算法中，从而使后者对权值 的调整由原来的两项增加为三项，从而提出一种新的学习算法---TWEBP算法。为了 验证本算法的优点，采用了三个例子，分别对异或问题、三分类问题和函数逼近问题 进行了实验，发现其收敛速度和逃离局部极小点的能力都优于传统算法。 关键词：多层感知器学习算法趋势外推均衡因子TWEBP 作 者：王之仓 指导教师：邓伟 Research on Multilayer Perceptron Learning Algorithm ABSTRACT Multilayer Perceptron is a sort of multilayer feed-forward single direct propagation network model. Because of its good nonlinear mapping ability, it is one of the basic models in the research and application of neural network at present, which has been widely applied to pattern recognition, image processing, function approximation, optimization computation, optional prediction, adaptation control and so on. Multilayer Perception trained with BP algorithm often has a low convergence speed as a natural drawback，because it is based on gradient descent method which is only local searching. When applied to an object function with many local minimums, it is not possible for BP algorithm to avoid being trapped in local minimum and to have a low converges speed. In a word, the research on BP algorithm has become very important for a long time. The purpose of this design task is to study the algorithms of Multilayer Perceptron, and a new BP algorithm is presented. Both BPWE algorithm (back-propagation by weight extrapolation) and TBP algorithm (a three-term back propagation algorithm) are based on weight value adjusted. Considered to add the proportional factor of the TBP algorithm into BPWE algorithm, it made the latter can adjust weight value by three terms too. A new BP algorithm，named TWEBP (the three-term weight extrapolation back propagation algorithm), is presented based on the two algo