问题描述
目前，我正在学习神经网络，并且正在尝试创建一个可以训练以识别手写字符的应用程序.对于这个问题，我使用前馈神经网络，当我训练它识别1个，2个或3个不同的字符时，它似乎可以工作.但是，当我尝试使网络学习3个以上的字符时，它将以40％-60％左右的错误百分比停滞不前.
Currently I'm learning about neural networks and I'm trying to create an application that can be trained to recognize handwritten characters.For this problem I use a feed-forward neural network and it seems to work when I train it to recognize 1, 2 or 3 different characters. But when I try to make the network learn more than 3 characters it will stagnate at a error percentage around the 40 - 60%.
我尝试使用多层和更少/更多的神经元，但是我似乎无法正确地做到这一点，现在我想知道前馈神经网络是否能够识别这么多信息.
I tried with multiple layers and less/more neurons but I can't seem to get it right, now I'm wondering if a feedforward neural network is capable of recognizing that much information.
一些统计数据:
网络类型:前馈神经网络
输入神经元:使用100(10 * 10)个网格绘制字符
Input neurons: 100 (a 10 * 10) grid is used to draw the characters
输出神经元:要重新定位的字符数
Output neurons: The amount of characters to regocnize
有人知道我的体系结构中可能存在的缺陷吗?输入神经元过多吗?前馈神经网络不能识别字符吗?
推荐答案
对于手写字符识别，您需要
For handwritten character recognition you need
- 许多训练示例(也许您应该使训练集失真)
- 输出层中的softmax激活功能
- 交叉熵误差函数
- 进行随机梯度下降训练
- 每层都有偏差
- many training examples (maybe you should create distortions of your training set)
- softmax activation function in the output layer
- cross entropy error function
- training with stochastic gradient descent
- a bias in each layer
一个好的测试问题是手写数字数据集 MNIST .以下是在此数据集上成功应用神经网络的论文:
A good test problem is the handwritten digit data set MNIST. Here are papers that successfully applied neural networks on this data set:
Y. LeCun，L.Bottou，Y.Bengio和P. Haffner:基于梯度的学习应用于文档识别， http://yann.lecun.com/exdb/publis/pdf/lecun-98.pdf
Y. LeCun, L. Bottou, Y. Bengio and P. Haffner: Gradient-Based Learning Applied to Document Recognition, http://yann.lecun.com/exdb/publis/pdf/lecun-98.pdf
Dan Claudiu Ciresan，Ueli Meier，Luca Maria Gambardella，Juergen Schmidhuber:手写数字识别的深层简单神经网络Excel， http://arxiv.org/abs/1003.0358
Dan Claudiu Ciresan, Ueli Meier, Luca Maria Gambardella, Juergen Schmidhuber: Deep Big Simple Neural Nets Excel on Handwritten Digit Recognition, http://arxiv.org/abs/1003.0358
我训练了具有784-200-50-10架构的MLP，并在测试集上获得了> 96％的准确度.
I trained an MLP with 784-200-50-10 architecture and got >96% accuracy on the test set.
这篇关于训练用于OCR的前馈神经网络的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！