课本P151 中的发射概率矩阵 表12-2的发射概率矩阵示例，是否有问题（均为隐藏状态S），代码12-2.py，调用apriori算法，计算了置信度，并没有进行HMM算法求参数矩阵，

Load file(../data/ADFA-LD/Training_Data_Master/UTD-0796.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0797.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0798.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0799.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0800.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0801.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0802.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0803.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0804.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0805.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0806.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0807.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0808.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0809.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0810.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0811.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0812.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0813.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0814.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0815.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0816.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0817.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0818.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0819.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0820.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0821.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0822.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0823.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0824.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0825.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0826.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0827.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0828.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0829.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0830.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0831.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0832.txt)
Load file(../data/ADFA-LD/Training_Data_Master/UTD-0833.txt)
Traceback (most recent call last):
File "D:\Users\Administrator\eclipsFaySon\secuit\code\8-2.py", line 76, in
score=model_selection.cross_val_score(logreg, x, y, n_jobs=-1, cv=10)
File "D:\ProgramData\Anaconda2\lib\site-packages\sklearn\model_selection_validation.py", line 342, in cross_val_score
pre_dispatch=pre_dispatch)
File "D:\ProgramData\Anaconda2\lib\site-packages\sklearn\model_selection_validation.py", line 206, in cross_validate
for train, test in cv.split(X, y, groups))
File "D:\ProgramData\Anaconda2\lib\site-packages\sklearn\externals\joblib\parallel.py", line 789, in call
self.retrieve()
File "D:\ProgramData\Anaconda2\lib\site-packages\sklearn\externals\joblib\parallel.py", line 740, in retrieve
raise exception
sklearn.externals.joblib.my_exceptions.JoblibValueError: JoblibValueError

Multiprocessing exception:
...........................................................................
D:\Users\Administrator\eclipsFaySon\secuit\code\8-2.py in ()
71 solver='sgd', verbose=10, tol=1e-4, random_state=1,
72 learning_rate_init=.1)
73
74 logreg = linear_model.LogisticRegression(C=1e5)
75
---> 76 score=model_selection.cross_val_score(logreg, x, y, n_jobs=-1, cv=10)
77 print np.mean(score)
78
79
80

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\model_selection_validation.py in cross_val_score(estimator=LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), X=array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64), y=[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...], groups=None, scoring=None, cv=10, n_jobs=-1, verbose=0, fit_params=None, pre_dispatch='2n_jobs')
337 cv_results = cross_validate(estimator=estimator, X=X, y=y, groups=groups,
338 scoring={'score': scorer}, cv=cv,
339 return_train_score=False,
340 n_jobs=n_jobs, verbose=verbose,
341 fit_params=fit_params,
--> 342 pre_dispatch=pre_dispatch)
pre_dispatch = '2n_jobs'
343 return cv_results['test_score']
344
345
346 def _fit_and_score(estimator, X, y, scorer, train, test, verbose,

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\model_selection_validation.py in cross_validate(estimator=LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), X=array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64), y=[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...], groups=None, scoring={'score': }, cv=StratifiedKFold(n_splits=10, random_state=None, shuffle=False), n_jobs=-1, verbose=0, fit_params=None, pre_dispatch='2*n_jobs', return_train_score=False)
201 scores = parallel(
202 delayed(_fit_and_score)(
203 clone(estimator), X, y, scorers, train, test, verbose, None,
204 fit_params, return_train_score=return_train_score,
205 return_times=True)
--> 206 for train, test in cv.split(X, y, groups))
cv.split = <bound method StratifiedKFold.split of Stratifie...d(n_splits=10, random_state=None, shuffle=False)>
X = array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64)
y = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...]
groups = None
207
208 if return_train_score:
209 train_scores, test_scores, fit_times, score_times = zip(*scores)
210 train_scores = _aggregate_score_dicts(train_scores)

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\externals\joblib\parallel.py in call(self=Parallel(n_jobs=-1), iterable=<generator object >)
784 if pre_dispatch == "all" or n_jobs == 1:
785 # The iterable was consumed all at once by the above for loop.
786 # No need to wait for async callbacks to trigger to
787 # consumption.
788 self._iterating = False
--> 789 self.retrieve()
self.retrieve = <bound method Parallel.retrieve of Parallel(n_jobs=-1)>
790 # Make sure that we get a last message telling us we are done
791 elapsed_time = time.time() - self._start_time
792 self._print('Done %3i out of %3i | elapsed: %s finished',
793 (len(self._output), len(self._output),

Sub-process traceback:

ValueError Wed Apr 3 08:53:56 2019
PID: 16256 Python 2.7.15: D:\ProgramData\Anaconda2\python.exe
...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\externals\joblib\parallel.py in call(self=<sklearn.externals.joblib.parallel.BatchedCalls object>)
126 def init(self, iterator_slice):
127 self.items = list(iterator_slice)
128 self._size = len(self.items)
129
130 def call(self):
--> 131 return [func(*args, **kwargs) for func, args, kwargs in self.items]
func =
args = (LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64), [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...], {'score': }, array([ 84, 85, 86, 87, 88, 89, 90, 91, ... 826, 827, 828, 829, 830, 831, 832], dtype=int64), array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1..., 77, 78, 79, 80, 81, 82, 83],
dtype=int64), 0, None, None)
kwargs = {'return_times': True, 'return_train_score': False}
self.items = [(, (LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64), [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...], {'score': }, array([ 84, 85, 86, 87, 88, 89, 90, 91, ... 826, 827, 828, 829, 830, 831, 832], dtype=int64), array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1..., 77, 78, 79, 80, 81, 82, 83],
dtype=int64), 0, None, None), {'return_times': True, 'return_train_score': False})]
132
133 def len(self):
134 return self._size
135

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\model_selection_validation.py in _fit_and_score(estimator=LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), X=array([[0, 0, 0, ..., 0, 0, 0],
[0, 0, 0,...0],
[0, 0, 0, ..., 0, 0, 0]], dtype=int64), y=[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...], scorer={'score': }, train=array([ 84, 85, 86, 87, 88, 89, 90, 91, ... 826, 827, 828, 829, 830, 831, 832], dtype=int64), test=array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 1..., 77, 78, 79, 80, 81, 82, 83],
dtype=int64), verbose=0, parameters=None, fit_params={}, return_train_score=False, return_parameters=False, return_n_test_samples=False, return_times=True, error_score='raise')
453
454 try:
455 if y_train is None:
456 estimator.fit(X_train, **fit_params)
457 else:
--> 458 estimator.fit(X_train, y_train, **fit_params)
estimator.fit = <bound method LogisticRegression.fit of Logistic...inear', tol=0.0001, verbose=0, warm_start=False)>
X_train = array([[ 0, 0, 0, ..., 0, 0, 0],
... 0, 0, 0, ..., 0, 0, 0]], dtype=int64)
y_train = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, ...]
fit_params = {}
459
460 except Exception as e:
461 # Note fit time as time until error
462 fit_time = time.time() - start_time

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\linear_model\logistic.py in fit(self=LogisticRegression(C=100000.0, class_weight=None...linear', tol=0.0001, verbose=0, warm_start=False), X=array([[ 0., 0., 0., ..., 0., 0., 0.]... [ 0., 0., 0., ..., 0., 0., 0.]]), y=array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0]), sample_weight=None)
1232 " = {}.".format(self.n_jobs))
1233 self.coef_, self.intercept_, n_iter_ = fit_liblinear(
1234 X, y, self.C, self.fit_intercept, self.intercept_scaling,
1235 self.class_weight, self.penalty, self.dual, self.verbose,
1236 self.max_iter, self.tol, self.random_state,
-> 1237 sample_weight=sample_weight)
sample_weight = None
1238 self.n_iter = np.array([n_iter_])
1239 return self
1240
1241 if self.solver in ['sag', 'saga']:

...........................................................................
D:\ProgramData\Anaconda2\lib\site-packages\sklearn\svm\base.py in fit_liblinear(X=array([[ 0., 0., 0., ..., 0., 0., 0.]... [ 0., 0., 0., ..., 0., 0., 0.]]), y=array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,...0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0]), C=100000.0, fit_intercept=True, intercept_scaling=1, class_weight=None, penalty='l2', dual=False, verbose=0, max_iter=100, tol=0.0001, random_state=None, multi_class='ovr', loss='logistic_regression', epsilon=0.1, sample_weight=None)
848 y_ind = enc.fit_transform(y)
849 classes = enc.classes_
850 if len(classes_) < 2:
851 raise ValueError("This solver needs samples of at least 2 classes"
852 " in the data, but the data contains only one"
--> 853 " class: %r" % classes_[0])
classes_ = array([0])
854
855 class_weight_ = compute_class_weight(class_weight, classes_, y)
856 else:
857 class_weight_ = np.empty(0, dtype=np.float64)

ValueError: This solver needs samples of at least 2 classes in the data, but the data contains only one class: 0

import matplotlib.pyplot as plt
from sklearn.datasets import fetch_mldata
from sklearn.neural_network import MLPClassifier

mnist = fetch_mldata("MNIST original")

rescale the data, use the traditional train/test split

X, y = mnist.data / 255., mnist.target
X_train, X_test = X[:60000], X[60000:]
y_train, y_test = y[:60000], y[60000:]

HTTPError: HTTP Error 404: Dataset 'mnist-original' not found on mldata.org.

只有50个用户的命令操作文件，标识文件在Schonlau的个人网站： http://www.schonlau.net/上面没有找到。

line 37:
webshell_bigram_vectorizer = CountVectorizer(ngram_range=(1, 1), decode_error="ignore",
token_pattern = r_token_pattern,min_df=1)

line 45:
wp_bigram_vectorizer = CountVectorizer(ngram_range=(2, 2), decode_error="ignore",
token_pattern = r_token_pattern,min_df=1,vocabulary=vocabulary)

在对黑样本的特征化过程中，用的是1-gram，提取出的 vocabulary 也是 1-gram

但是在白样本的特征化中，使用的仍然是 2-gram，提取出的 字符串 不可能出现在 vocabulary 中啊..
白样本的特征向量全都为 [0 0 0 ... 0 0 0]

Warning (from warnings module):
File "C:\Users\lenvov\AppData\Local\Programs\Python\Python38\lib\site-packages\sklearn\model_selection_split.py", line 665
warnings.warn(("The least populated class in y has only %d"
UserWarning: The least populated class in y has only 2 members, which is less than n_splits=10.
[1. 1. 1. 1. 1. 0.77777778

1.     1.         0.88888889 0.88888889]
   

5-4.py运行结果如上面所示，10折交叉验证存在问题，与样本有关

requirements.txt better.
thx a lot :)

兜哥您好，本代码当N=100时python3.6运行出来的结果是100，而您书上给出的是83.3333。当我把N改成90时，结果就是83.333了。对于这问题希望您解释一下，是否是代码里有些问题呢？

兜哥，你好

WEB安全机器学习入门P78里面有个“统计最频繁使用的前50个命令和最不频繁50个命令”对应的代码如下：

fdist = FreqDist(dist).keys()
dist_max=set(fdist[0:50])
dist_min = set(fdist[-50:])

我看了官网对FreqDist的解释，该函数功能是对词频进行统计，例如这里如果使用临时变量tmp进行遍历即可得到如下结果：

tmp  = FreqDist(dist)
for key in tmp:
    print key+" : "+str(tmp[key])

gs : 4
tset : 1
basename : 616
uname : 443
touch : 3
... ...
所以这里已经是去重后的数据，但使用fdist[0:50]和fdist[-50:]并不能得到最频繁使用的前50个命令和最不频繁50个命令，比如fdist[0:10]得到的就是字典前10条。

求指导！谢谢了！

2017年11月第一版第三次印刷的书从第126直接跳到143了

里面的样本有很多不懂，可以请教一下吗？建个群之类的？ 微信facial58

您好，方便给发一份代码吗，下载了三天，实在是下载不下来。
我的邮箱是：[email protected]。
很感谢您能公布代码

P132，倒数第四行，原文："from sklearn.cluster import K-Means"，纠正为："from sklearn.cluster import KMeans"

很多脚本使用10折交叉验证, sklearn 0.18以上已经弃用, 改用sklearn.model_selection.cross_val_score instead

The second one is the cloaked images that I used your algorithm in th == 0.007. Is there really exist the reslut like you say?

兜哥，你好
我在学习过程中发现的书中的一些问题，如果我的勘误不对的话，还请兜哥原谅

P12 图2-1 主流基于Python的机器学习库 横纵坐标单词拼写错误 分别为Commits、Contributors
P14 >>>a = array([1, 2, 3]) 应为 a = np.array([1, 2, 3])
P15 >>>c.shape(2,3,2,3) 应为 c.reshape(2,3,2,3)
P37 应修改为
TensorFlow中有类似实现：

from sklearn.feature_extraction.text import CountVectorizer
from tensorflow.contrib import learn
MAX_DOCUMENT_LENGTH=100

7-6.py中使用mnist.pkl.gz，在python3.6中读取其中的值会报错UnicodeDecodeError: 'ascii' codec can't decode byte 0x90 in position 614: ordinal not in range(128)

首先，在这里_y应该改成y_吧？
然后，y_对应的应该是mnist.train.labels（注：mnist.train.labels是一个 [60000, 10] 的数字矩阵，用来表示训练数据集的标签）的第二个维度对应的one-hot编码之后的值，而不是整个系统的输出吧？
而整个系统的输出应该是y吧？

同时x对应的是mnist.train.images的第二个维度对应的值吧？（mnist.train.images 是一个形状为 [60000, 784] 的张量，第一个维度数字用来索引图片，第二个维度数字用来索引每张图片中的像素点。）
而整个系统的输入应该是x和y_共同构成的吧？

P134，最后一行，原文："plt.show()”缩进有问题，应该与for对齐。
P139，第2行，原文：“那么称这个事件A。为k项集事件A” 应改为 “那么称这个事件A为k项集事件A”
P151， 表12-2，发射概率应该是隐藏状态下输出某个观察状态的概率，表12-2应为S1-S4向ACNT的发射概率矩阵。
P153, 第一行，原文“参数异常就输入第三种” 应改为 “参数异常就属于第三种”
P160， 第12行，原文： “if domain>=MIN_LEN:” 应改为"if len(domain)>=MIN:"
P160, 倒数第13行，原文： ‘if domain >= MIN_LEN:’ 应改为 “if len(domain)>=MIN:”
P165 , 第1行，原文：”ne04j启动“和 第2行，默认密码为ne04j/ne04j， 应修改为”neo4j启动“和“neo4j/neo4j”
P185， 第3行，原文：“初步怀疑邮箱file1和file1为同一黑产控制的后门文件” 应修改为“初步怀疑file1和file2为同一黑产控制的后门文件”
P190， 倒数第5行 图14-6，原文： “24X24的图片”，应该为：“28X28的图片”
P201， 倒数第13行和第14行，原文“y” 应该为“y”
P203， 倒数第10行和第11行，原文“y” 应该为“y”

下面这个问题需要解答一下：
P143， 倒数第二行和倒数第一行，原文：“开始节点的值表示这个项集的支持度”，是不是应该说开始节点的值表示的是这个项集的出现次数呢？ 因为支持度是一个比例值，应该是0-1之间的？ 倒数第一行，“项集{z}的支持度为5.。。。”应该是说出现次数为5吧？
P144页， 第5行，原文：“最小支持度设为3”，支持度是个比例值吗？兜哥求解答，O(∩_∩)O谢谢。

兜哥15-3.py的代码跟书本上的代码块严重不符啊更新一下代码呗

你好，最近开始学习您的web安全机器学习一书，请问MasqueradeDat/label.txt这个文件是怎么生成的呢？看到您的代码中应用了这个文件，但是我看原始数据集中并没有这个文件，所以想问一下这个文件是做什么用的，是如何生成的呢？