neural_network.convolution_neural_network

          • – - - - - - - - - - - - - - - - - - - - - - -

Name - - CNN - Convolution Neural Network For Photo Recognizing Goal - - Recognize Handwriting Word Photo Detail: Total 5 layers neural network

  • Convolution layer

  • Pooling layer

  • Input layer layer of BP

  • Hidden layer of BP

  • Output layer of BP

Author: Stephen Lee Github: 245885195@qq.com Date: 2017.9.20 - - - - - – - - - - - - - - - - - - - - - - - - - - - -

Classes

CNN

Module Contents

class neural_network.convolution_neural_network.CNN(conv1_get, size_p1, bp_num1, bp_num2, bp_num3, rate_w=0.2, rate_t=0.2)
_calculate_gradient_from_pool(out_map, pd_pool, num_map, size_map, size_pooling)

calculate the gradient from the data slice of pool layer pd_pool: list of matrix out_map: the shape of data slice(size_map*size_map) return: pd_all: list of matrix, [num, size_map, size_map]

_expand(data)
_expand_mat(data_mat)
convolute(data, convs, w_convs, thre_convs, conv_step)
convolution(data)
do_round(x)
pooling(featuremaps, size_pooling, pooling_type='average_pool')
predict(datas_test)
classmethod read_model(model_path)
save_model(save_path)
sig(x)
train(patterns, datas_train, datas_teach, n_repeat, error_accuracy, draw_e=bool)
conv1
num_bp1
num_bp2
num_bp3
rate_thre = 0.2
rate_weight = 0.2
size_pooling1
step_conv1
thre_bp2
thre_bp3
thre_conv1
vji
w_conv1
wkj