1
目录
分类算子 ............................................................................................................................................................................................. 3
7.6 Support Vector Machines ........................................................................................................................................ 3
add_class_train_data_svm ...................................................................................................................................... 3
add_sample_class_svm ............................................................................................................................................... 3
classify_class_svm .................................................................................................................................................... 3
clear_class_svm ........................................................................................................................................................... 4
clear_samples_class_svm ........................................................................................................................................ 4
create_class_svm ........................................................................................................................................................ 4
evaluate_class_svm .................................................................................................................................................... 7
get_class_train_data_svm ...................................................................................................................................... 8
get_params_class_svm ............................................................................................................................................... 8
get_prep_info_class_svm ........................................................................................................................................ 8
get_sample_class_svm ............................................................................................................................................... 8
get_sample_num_class_svm ...................................................................................................................................... 8
get_support_vector_class_svm ............................................................................................................................. 8
get_support_vector_num_class_svm .................................................................................................................... 8
read_class_svm ............................................................................................................................................................. 8
read_samples_class_svm ........................................................................................................................................... 8
reduce_class_svm ........................................................................................................................................................ 9
select_feature_set_svm ........................................................................................................................................... 9
train_class_svm ........................................................................................................................................................... 9
write_class_svm ........................................................................................................................................................... 9
write_samples_class_svm ........................................................................................................................................ 9
1 简介 ................................................................................................................................................................................................ 10
2 一个例程 ....................................................................................................................................................................................... 11
3 分类的理论背景 ......................................................................................................................................................................... 13
3.1 通常的分类 .................................................................................................................................................................... 14
3.2 Euclidean and Hyperbox Classifiers ............................................................................................................. 15
3.3 Multi-Layer Perceptrons (MLP) ............................................................................................................... 16
3.4 Support-Vector Machines (SVM) ............................................................................................................... 16
3.5 Gaussian Mixture Models (GMM) ............................................................................................................... 18
3.6 K-Nearest Neighbors (k-NN) ...................................................................................................................... 18
3.7 Convolutional Neural Networks (CNNs) ............................................................................................... 18
4 Decisions to Make .................................................................................................................................................................. 18
4.1 选择一个合适的分类方法 ......................................................................................................................................... 18
4.2 选择合适的特征 ........................................................................................................................................................... 19
4.3 选择合适的训练样本 .................................................................................................................................................. 20
5 分类的一般特征 ......................................................................................................................................................................... 20
5.1 一般方法 ......................................................................................................................................................................... 21
5.2 有关的算子(概览) .................................................................................................................................................. 23
5.2.1 基本步骤 ........................................................................................................................................................... 23
5.2.2 高级步骤 ........................................................................................................................................................... 25
5.4 Parameter Setting for SVM .................................................................................................................................. 26
5.4.1 调整 create_class_svm .............................................................................................................................. 26