616 lines
42 KiB
Plaintext
616 lines
42 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 1,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"import os\n",
|
|
"import json\n",
|
|
"import pandas as pd\n",
|
|
"import numpy as np\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 2,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"ciper_suits = {\n",
|
|
" '1305':0,\n",
|
|
" 'C030':1,\n",
|
|
"\t'C02C':2,\n",
|
|
"\t'C028':3,\n",
|
|
"\t'C024':4,\n",
|
|
"\t'C014':5,\n",
|
|
"\t'C00A':6,\n",
|
|
"\t'00A5':7,\n",
|
|
"\t'00A3':8,\n",
|
|
"\t'00A1':9,\n",
|
|
"\t'009F':10,\n",
|
|
"\t'006B':11,\n",
|
|
"\t'006A':12,\n",
|
|
"\t'0069':13,\n",
|
|
"\t'0068':14,\n",
|
|
"\t'0039':15,\n",
|
|
"\t'0038':16,\n",
|
|
"\t'0037':17,\n",
|
|
"\t'0036':18,\n",
|
|
"\t'0088':19,\n",
|
|
"\t'0087':20,\n",
|
|
"\t'0086':21,\n",
|
|
"\t'0085':22,\n",
|
|
"\t'C019':23,\n",
|
|
"\t'00A7':24,\n",
|
|
"\t'006D':25,\n",
|
|
"\t'003A':26,\n",
|
|
"\t'0089':27,\n",
|
|
"\t'C032':28,\n",
|
|
"\t'C02E':29,\n",
|
|
"\t'C02A':30,\n",
|
|
"\t'C026':31,\n",
|
|
"\t'C00F':32,\n",
|
|
"\t'C005':33,\n",
|
|
"\t'009D':34,\n",
|
|
"\t'003D':35,\n",
|
|
"\t'0035':36,\n",
|
|
"\t'0084':37,\n",
|
|
"\t'008D':38,\n",
|
|
"\t'C02F':39,\n",
|
|
"\t'C02B':40,\n",
|
|
"\t'C027':41,\n",
|
|
"\t'C023':42,\n",
|
|
"\t'C013':43,\n",
|
|
"\t'C009':44,\n",
|
|
"\t'00A4':45,\n",
|
|
"\t'00A2':46,\n",
|
|
"\t'00A0':47,\n",
|
|
"\t'009E':48,\n",
|
|
"\t'0067':49,\n",
|
|
"\t'0040':50,\n",
|
|
"\t'003F':51,\n",
|
|
"\t'003E':52,\n",
|
|
"\t'0033':53,\n",
|
|
"\t'0032':54,\n",
|
|
"\t'0031':55,\n",
|
|
"\t'0030':56,\n",
|
|
"\t'009A':57,\n",
|
|
"\t'0099':58,\n",
|
|
"\t'0098':59,\n",
|
|
"\t'0097':60,\n",
|
|
"\t'0045':61,\n",
|
|
"\t'0044':62,\n",
|
|
"\t'0043':63,\n",
|
|
"\t'0042':64,\n",
|
|
"\t'C018':65,\n",
|
|
"\t'00A6':66,\n",
|
|
"\t'006C':67,\n",
|
|
"\t'0034':68,\n",
|
|
"\t'009B':69,\n",
|
|
"\t'0046':70,\n",
|
|
"\t'C031':71,\n",
|
|
"\t'C02D':72,\n",
|
|
"\t'C029':73,\n",
|
|
"\t'C025':74,\n",
|
|
"\t'C00E':75,\n",
|
|
"\t'C004':76,\n",
|
|
"\t'009C':77,\n",
|
|
"\t'003C':78,\n",
|
|
"\t'002F':79,\n",
|
|
"\t'0096':80,\n",
|
|
"\t'0041':81,\n",
|
|
"\t'008C':82,\n",
|
|
"\t'C012':83,\n",
|
|
"\t'C008':84,\n",
|
|
"\t'0016':85,\n",
|
|
"\t'0013':86,\n",
|
|
"\t'0010':87,\n",
|
|
"\t'000D':88,\n",
|
|
"\t'C017':89,\n",
|
|
"\t'001B':90,\n",
|
|
"\t'C00D':91,\n",
|
|
"\t'C003':92,\n",
|
|
"\t'000A':93,\n",
|
|
"\t'0007':94,\n",
|
|
"\t'008B':95,\n",
|
|
"\t'0021':96,\n",
|
|
"\t'001F':97,\n",
|
|
"\t'0025':98,\n",
|
|
"\t'0023':99,\n",
|
|
"\t'C011':100,\n",
|
|
"\t'C007':101,\n",
|
|
"\t'C016':102,\n",
|
|
"\t'0018':103,\n",
|
|
"\t'C00C':104,\n",
|
|
"\t'C002':105,\n",
|
|
"\t'0005':106,\n",
|
|
"\t'0004':107,\n",
|
|
"\t'008A':108,\n",
|
|
"\t'0020':109,\n",
|
|
"\t'0024':110,\n",
|
|
"\t'C010':111,\n",
|
|
"\t'C006':112,\n",
|
|
"\t'C015':113,\n",
|
|
"\t'C00B':114,\n",
|
|
"\t'C001':115,\n",
|
|
"\t'003B':116,\n",
|
|
"\t'0002':117,\n",
|
|
"\t'0001':118,\n",
|
|
" '1301':119,\n",
|
|
"\t'1302':120,\n",
|
|
"\t'1303':121,\n",
|
|
"\t'1304':122\n",
|
|
"}"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 3,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"extensions = { \n",
|
|
" 0:0, \n",
|
|
" 1:1, \n",
|
|
" 2:2, \n",
|
|
" 3:3, \n",
|
|
" 4:4, \n",
|
|
" 5:5, \n",
|
|
" 6:6, \n",
|
|
" 7:7, \n",
|
|
" 8:8, \n",
|
|
" 9:9, \n",
|
|
" 10:10, \n",
|
|
" 11:11, \n",
|
|
" 12:12, \n",
|
|
" 13:13, \n",
|
|
" 14:14, \n",
|
|
" 15:15, \n",
|
|
" 16:16, \n",
|
|
" 17:17, \n",
|
|
" 18:18, \n",
|
|
" 19:19, \n",
|
|
" 20:20, \n",
|
|
" 21:21, \n",
|
|
" 22:22, \n",
|
|
" 23:23, \n",
|
|
" 24:24, \n",
|
|
" 25:25, \n",
|
|
" 26:26, \n",
|
|
" 27:27, \n",
|
|
" 28:28, \n",
|
|
" 29:29, \n",
|
|
" 30:30, \n",
|
|
" 31:31, \n",
|
|
" 35:32, \n",
|
|
" 65281:33 \n",
|
|
"}"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 4,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"date = '2019-12-20_21'\n",
|
|
"root_dir = '/Users/Leo/Documents/github/GradProj/'\n",
|
|
"example_label_file = root_dir + 'DataSet/result/' + date + '/stream_tag.txt'\n",
|
|
"example_label_df = pd.read_table(example_label_file, sep='\\s+', header=None)\n",
|
|
"example_label = {tuple(example_label_df.iloc[i,0:4].values):example_label_df.iloc[i,4] for i in example_label_df.index}\n",
|
|
"example_json_file = root_dir + 'DataSet/result/' + date + '/stream_stat.txt'\n",
|
|
"example_json_f = open(example_json_file, 'r')\n",
|
|
"result_data = list()\n",
|
|
"result_label = list()\n",
|
|
"i = 0\n",
|
|
"for line in example_json_f.readlines():\n",
|
|
" example_json = json.loads(line)\n",
|
|
" #标签\n",
|
|
" try:\n",
|
|
" flow_key = (example_json['sip'], example_json['sport'], example_json['dip'], example_json['dport'])\n",
|
|
" result_label.append(example_label[flow_key])\n",
|
|
" except Exception:\n",
|
|
" continue\n",
|
|
" \n",
|
|
" #统计特征\n",
|
|
" packets = example_json['packets']\n",
|
|
" c2s_packets_bytes = list()\n",
|
|
" s2c_packets_bytes = list()\n",
|
|
" c2s_packets_intervals = list()\n",
|
|
" s2c_packets_intervals = list()\n",
|
|
" for packet in packets:\n",
|
|
" if packet['dir'] == 1:\n",
|
|
" c2s_packets_bytes.append(packet['bytes'])\n",
|
|
" c2s_packets_intervals.append(packet['interval'])\n",
|
|
" elif packet['dir'] == 2:\n",
|
|
" s2c_packets_bytes.append(packet['bytes'])\n",
|
|
" s2c_packets_intervals.append(packet['interval'])\n",
|
|
" c2s_bytes = example_json['c2s_bytes']\n",
|
|
" s2c_bytes = example_json['s2c_bytes']\n",
|
|
" c2s_pkts = example_json['c2s_pkts']\n",
|
|
" s2c_pkts = example_json['s2c_pkts']\n",
|
|
" duration = example_json['duration']\n",
|
|
" c2s_packets_bytes_mean = 0\n",
|
|
" c2s_packets_bytes_median = 0\n",
|
|
" c2s_packets_bytes_std = 0\n",
|
|
" c2s_packets_bytes_max = 0\n",
|
|
" c2s_packets_bytes_min = 0\n",
|
|
"\n",
|
|
" c2s_packets_intervals_mean = 0\n",
|
|
" c2s_packets_intervals_median = 0\n",
|
|
" c2s_packets_intervals_std = 0\n",
|
|
" c2s_packets_intervals_max = 0\n",
|
|
" c2s_packets_intervals_min = 0\n",
|
|
"\n",
|
|
" s2c_packets_bytes_mean = 0\n",
|
|
" s2c_packets_bytes_median = 0\n",
|
|
" s2c_packets_bytes_std = 0\n",
|
|
" s2c_packets_bytes_max = 0\n",
|
|
" s2c_packets_bytes_min = 0\n",
|
|
"\n",
|
|
" s2c_packets_intervals_mean = 0\n",
|
|
" s2c_packets_intervals_median = 0\n",
|
|
" s2c_packets_intervals_std = 0\n",
|
|
" s2c_packets_intervals_max = 0\n",
|
|
" s2c_packets_intervals_min = 0\n",
|
|
" \n",
|
|
" if c2s_bytes > 0:\n",
|
|
" c2s_packets_bytes_mean = np.mean(c2s_packets_bytes)\n",
|
|
" c2s_packets_bytes_median = np.median(c2s_packets_bytes)\n",
|
|
" c2s_packets_bytes_std = np.std(c2s_packets_bytes)\n",
|
|
" c2s_packets_bytes_max = np.max(c2s_packets_bytes)\n",
|
|
" c2s_packets_bytes_min = np.min(c2s_packets_bytes)\n",
|
|
"\n",
|
|
" c2s_packets_intervals_mean = np.mean(c2s_packets_intervals)\n",
|
|
" c2s_packets_intervals_median = np.median(c2s_packets_intervals)\n",
|
|
" c2s_packets_intervals_std = np.std(c2s_packets_intervals)\n",
|
|
" c2s_packets_intervals_max = np.max(c2s_packets_intervals)\n",
|
|
" c2s_packets_intervals_min = np.min(c2s_packets_intervals)\n",
|
|
" \n",
|
|
" if s2c_bytes > 0:\n",
|
|
" s2c_packets_bytes_mean = np.mean(s2c_packets_bytes)\n",
|
|
" s2c_packets_bytes_median = np.median(s2c_packets_bytes)\n",
|
|
" s2c_packets_bytes_std = np.std(s2c_packets_bytes)\n",
|
|
" s2c_packets_bytes_max = np.max(s2c_packets_bytes)\n",
|
|
" s2c_packets_bytes_min = np.min(s2c_packets_bytes)\n",
|
|
"\n",
|
|
" s2c_packets_intervals_mean = np.mean(s2c_packets_intervals)\n",
|
|
" s2c_packets_intervals_median = np.median(s2c_packets_intervals)\n",
|
|
" s2c_packets_intervals_std = np.std(s2c_packets_intervals)\n",
|
|
" s2c_packets_intervals_max = np.max(s2c_packets_intervals)\n",
|
|
" s2c_packets_intervals_min = np.min(s2c_packets_intervals)\n",
|
|
"\n",
|
|
" #tls\n",
|
|
" tls = example_json['tls']\n",
|
|
" extensions_list = tls['extensions_list']\n",
|
|
" #print(extensions_list)\n",
|
|
" ciphers = tls['cipher_suites']\n",
|
|
" #print(ciphers)\n",
|
|
" extensions_arr = np.zeros(34, dtype=np.uint8)\n",
|
|
" cipher_suits_arr = np.zeros(123, dtype=np.uint8)\n",
|
|
" for extension in extensions_list:\n",
|
|
" try:\n",
|
|
" extensions_arr[extensions[extension]]=1\n",
|
|
" except Exception:\n",
|
|
" pass\n",
|
|
" for cipher in ciphers:\n",
|
|
" try:\n",
|
|
" cipher = cipher.upper()\n",
|
|
" cipher_suits_arr[ciper_suits[cipher]]=1\n",
|
|
" except Exception:\n",
|
|
" pass\n",
|
|
" result = [c2s_bytes, c2s_pkts, s2c_bytes, s2c_pkts, duration, c2s_packets_bytes_mean, c2s_packets_bytes_median, c2s_packets_bytes_std,\\\n",
|
|
" c2s_packets_bytes_max, c2s_packets_bytes_min, c2s_packets_intervals_mean, c2s_packets_intervals_median, c2s_packets_intervals_std,\\\n",
|
|
" c2s_packets_intervals_max, c2s_packets_intervals_min, s2c_packets_bytes_mean, s2c_packets_bytes_median, s2c_packets_bytes_std,\\\n",
|
|
" s2c_packets_bytes_max, s2c_packets_bytes_min, s2c_packets_intervals_mean, s2c_packets_intervals_median, s2c_packets_intervals_std,\\\n",
|
|
" s2c_packets_intervals_max, s2c_packets_intervals_min]\n",
|
|
" result += list(cipher_suits_arr)\n",
|
|
" result += list(extensions_arr)\n",
|
|
" result_data.append(result)\n",
|
|
" i += 1\n",
|
|
"extensions_head = list()\n",
|
|
"for i in range(len(extensions)):\n",
|
|
" extensions_head.append('extension'+str(i))\n",
|
|
"cipher_head = ['cipher'+str(i) for i in range(len(ciper_suits))]\n",
|
|
"base_head = ['c2s_bytes', 'c2s_pkts', 's2c_bytes', 's2c_pkts', 'duration', 'c2s_packets_bytes_mean', 'c2s_packets_bytes_median', 'c2s_packets_bytes_std',\\\n",
|
|
" 'c2s_packets_bytes_max', 'c2s_packets_bytes_min', 'c2s_packets_intervals_mean', 'c2s_packets_intervals_median', 'c2s_packets_intervals_std',\\\n",
|
|
" 'c2s_packets_intervals_max', 'c2s_packets_intervals_min', 's2c_packets_bytes_mean', 's2c_packets_bytes_median', 's2c_packets_bytes_std',\\\n",
|
|
" 's2c_packets_bytes_max', 's2c_packets_bytes_min', 's2c_packets_intervals_mean', 's2c_packets_intervals_median', 's2c_packets_intervals_std',\\\n",
|
|
" 's2c_packets_intervals_max', 's2c_packets_intervals_min']\n",
|
|
"header = base_head+cipher_head+extensions_head\n",
|
|
"result_df = pd.DataFrame(result_data, columns=header)\n",
|
|
"result_df['label'] = np.array(result_label)\n",
|
|
"example_csv_file = root_dir + 'Experiment/StatFeature/CsvFile/' + date + '/examples.csv'\n",
|
|
"result_df.to_csv(example_csv_file, index=False)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 5,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": [
|
|
"%matplotlib inline\n",
|
|
"import os\n",
|
|
"import numpy as np\n",
|
|
"import pandas as pd\n",
|
|
"import matplotlib.pyplot as plt"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 6,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"hupu: 489846\n",
|
|
"weibo: 897897\n",
|
|
"douyin: 158497\n",
|
|
"toutiao: 213989\n",
|
|
"zhihu: 968036\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# 统计每个app的包数\n",
|
|
"date = '2019-12-20_21'\n",
|
|
"root_dir = '/Users/Leo/Documents/github/GradProj/'\n",
|
|
"exmaples_file = root_dir + 'Experiment/StatFeature/CsvFile/' + date + '/examples.csv'\n",
|
|
"app2pktsDict = dict()\n",
|
|
"with open(exmaples_file) as f:\n",
|
|
" lines = f.readlines()\n",
|
|
" i = 0\n",
|
|
" for line in lines:\n",
|
|
" if i == 0:\n",
|
|
" i += 1\n",
|
|
" continue;\n",
|
|
" line = line.split(',')\n",
|
|
" pkts = int(line[1]) + int(line[3])\n",
|
|
" appName = line[-1]\n",
|
|
" if appName not in app2pktsDict.keys():\n",
|
|
" app2pktsDict[appName] = 0\n",
|
|
" app2pktsDict[appName] += pkts \n",
|
|
"for appName, pkts in app2pktsDict.items():\n",
|
|
" appName = appName[:-1]\n",
|
|
" print(appName + ': ', pkts)\n",
|
|
" "
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 7,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"data": {
|
|
"text/plain": [
|
|
"zhihu 6403\n",
|
|
"weibo 5487\n",
|
|
"douyin 3964\n",
|
|
"hupu 2304\n",
|
|
"toutiao 520\n",
|
|
"Name: label, dtype: int64"
|
|
]
|
|
},
|
|
"execution_count": 7,
|
|
"metadata": {},
|
|
"output_type": "execute_result"
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAEVCAYAAADpbDJPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAEnq1P8HV9/axntrFV8AAAAASUVORK5CYII=\n",
|
|
"text/plain": [
|
|
"<matplotlib.figure.Figure at 0x10ac1f320>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"examples_df = pd.read_csv(exmaples_file)\n",
|
|
"class_counts = examples_df['label'].value_counts()\n",
|
|
"class_counts.plot.bar()\n",
|
|
"class_counts"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 11,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"183\n",
|
|
" precision recall f1\n",
|
|
"LogisticRegression 0.773019 0.773019 0.773019\n",
|
|
"SVM 0.841328 0.841328 0.841328\n",
|
|
"GaussianNB 0.730193 0.730193 0.730193\n",
|
|
"tree 0.984154 0.984154 0.984154\n",
|
|
"RandomForest 0.988651 0.988651 0.988651\n"
|
|
]
|
|
},
|
|
{
|
|
"data": {
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAFcCAYAAAAzq/4LAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
|
|
"text/plain": [
|
|
"<matplotlib.figure.Figure at 0x112e0e550>"
|
|
]
|
|
},
|
|
"metadata": {},
|
|
"output_type": "display_data"
|
|
}
|
|
],
|
|
"source": [
|
|
"from sklearn.linear_model import LogisticRegression\n",
|
|
"from sklearn.svm import SVC\n",
|
|
"from sklearn.naive_bayes import GaussianNB\n",
|
|
"from sklearn import tree\n",
|
|
"from sklearn.ensemble import RandomForestClassifier\n",
|
|
"from sklearn.metrics import f1_score,recall_score,precision_score\n",
|
|
"import random\n",
|
|
"examples = examples_df.values.copy()\n",
|
|
"print(len(examples[0]))\n",
|
|
"#只取25个流统计特征\n",
|
|
"examples = np.c_[examples[:,:25].copy(),examples[:,-1].copy()]\n",
|
|
"#print(examples)\n",
|
|
"score_df = pd.DataFrame(np.zeros((5,3)),index = ['LogisticRegression', 'SVM', 'GaussianNB', 'tree', 'RandomForest'], \\\n",
|
|
" columns = ['precision', 'recall', 'f1'])\n",
|
|
"#def a():\n",
|
|
"\n",
|
|
"f1_score_list = list()\n",
|
|
"recall_score_list = list()\n",
|
|
"precision_score_list = list()\n",
|
|
"for i in range(1):\n",
|
|
" np.random.shuffle(examples)\n",
|
|
" examples_train = examples[:int(len(examples)*0.75)]\n",
|
|
" examples_test = examples[int(len(examples)*0.75):]\n",
|
|
" x_train = examples_train[:,0:-1]\n",
|
|
" y_train = examples_train[:,-1]\n",
|
|
" x_test = examples_test[:,0:-1]\n",
|
|
" y_test = examples_test[:,-1]\n",
|
|
" classifer = LogisticRegression()\n",
|
|
" classifer.fit(x_train, y_train)\n",
|
|
" y_pred = classifer.predict(x_test)\n",
|
|
" f1_score_list.append(f1_score(y_test, y_pred, average='micro'))\n",
|
|
" recall_score_list.append(recall_score(y_test, y_pred, average='micro'))\n",
|
|
" precision_score_list.append(precision_score(y_test, y_pred, average='micro'))\n",
|
|
"scores = [np.mean(precision_score_list), np.mean(recall_score_list), np.mean(f1_score_list)]\n",
|
|
"score_df.loc['LogisticRegression'] = scores\n",
|
|
"\n",
|
|
"f1_score_list = list()\n",
|
|
"recall_score_list = list()\n",
|
|
"precision_score_list = list()\n",
|
|
"for i in range(1):\n",
|
|
" #np.random.shuffle(examples)\n",
|
|
" examples_train = examples[:int(len(examples)*0.75)]\n",
|
|
" examples_test = examples[int(len(examples)*0.75):]\n",
|
|
" x_train = examples_train[:,0:-1]\n",
|
|
" y_train = examples_train[:,-1]\n",
|
|
" x_test = examples_test[:,0:-1]\n",
|
|
" y_test = examples_test[:,-1]\n",
|
|
" classifer = SVC()\n",
|
|
" classifer.fit(x_train, y_train)\n",
|
|
" y_pred = classifer.predict(x_test)\n",
|
|
" f1_score_list.append(f1_score(y_test, y_pred, average='micro'))\n",
|
|
" recall_score_list.append(recall_score(y_test, y_pred, average='micro'))\n",
|
|
" precision_score_list.append(precision_score(y_test, y_pred, average='micro'))\n",
|
|
"scores = [np.mean(precision_score_list), np.mean(recall_score_list), np.mean(f1_score_list)]\n",
|
|
"score_df.loc['SVM'] = scores\n",
|
|
"\n",
|
|
"f1_score_list = list()\n",
|
|
"recall_score_list = list()\n",
|
|
"precision_score_list = list()\n",
|
|
"for i in range(1):\n",
|
|
" #np.random.shuffle(examples)\n",
|
|
" examples_train = examples[:int(len(examples)*0.75)]\n",
|
|
" examples_test = examples[int(len(examples)*0.75):]\n",
|
|
" x_train = examples_train[:,0:-1]\n",
|
|
" y_train = examples_train[:,-1]\n",
|
|
" x_test = examples_test[:,0:-1]\n",
|
|
" y_test = examples_test[:,-1]\n",
|
|
" classifer = GaussianNB()\n",
|
|
" classifer.fit(x_train, y_train)\n",
|
|
" y_pred = classifer.predict(x_test)\n",
|
|
" f1_score_list.append(f1_score(y_test, y_pred, average='micro'))\n",
|
|
" recall_score_list.append(recall_score(y_test, y_pred, average='micro'))\n",
|
|
" precision_score_list.append(precision_score(y_test, y_pred, average='micro'))\n",
|
|
"scores = [np.mean(precision_score_list), np.mean(recall_score_list), np.mean(f1_score_list)]\n",
|
|
"score_df.loc['GaussianNB'] = scores\n",
|
|
"\n",
|
|
"f1_score_list = list()\n",
|
|
"recall_score_list = list()\n",
|
|
"precision_score_list = list()\n",
|
|
"for i in range(1):\n",
|
|
" #np.random.shuffle(examples)\n",
|
|
" examples_train = examples[:int(len(examples)*0.75)]\n",
|
|
" examples_test = examples[int(len(examples)*0.75):]\n",
|
|
" x_train = examples_train[:,0:-1]\n",
|
|
" y_train = examples_train[:,-1]\n",
|
|
" x_test = examples_test[:,0:-1]\n",
|
|
" y_test = examples_test[:,-1]\n",
|
|
" classifer = tree.DecisionTreeClassifier()\n",
|
|
" classifer.fit(x_train, y_train)\n",
|
|
" y_pred = classifer.predict(x_test)\n",
|
|
" f1_score_list.append(f1_score(y_test, y_pred, average='micro'))\n",
|
|
" recall_score_list.append(recall_score(y_test, y_pred, average='micro'))\n",
|
|
" precision_score_list.append(precision_score(y_test, y_pred, average='micro'))\n",
|
|
"scores = [np.mean(precision_score_list), np.mean(recall_score_list), np.mean(f1_score_list)]\n",
|
|
"score_df.loc['tree'] = scores\n",
|
|
"\n",
|
|
"f1_score_list = list()\n",
|
|
"recall_score_list = list()\n",
|
|
"precision_score_list = list()\n",
|
|
"for i in range(1):\n",
|
|
" #np.random.shuffle(examples)\n",
|
|
" examples_train = examples[:int(len(examples)*0.75)]\n",
|
|
" examples_test = examples[int(len(examples)*0.75):]\n",
|
|
" x_train = examples_train[:,0:-1]\n",
|
|
" y_train = examples_train[:,-1]\n",
|
|
" x_test = examples_test[:,0:-1]\n",
|
|
" y_test = examples_test[:,-1]\n",
|
|
" classifer = RandomForestClassifier()\n",
|
|
" classifer.fit(x_train, y_train)\n",
|
|
" y_pred = classifer.predict(x_test)\n",
|
|
" f1_score_list.append(f1_score(y_test, y_pred, average='micro'))\n",
|
|
" recall_score_list.append(recall_score(y_test, y_pred, average='micro'))\n",
|
|
" precision_score_list.append(precision_score(y_test, y_pred, average='micro'))\n",
|
|
"scores = [np.mean(precision_score_list), np.mean(recall_score_list), np.mean(f1_score_list)]\n",
|
|
"score_df.loc['RandomForest'] = scores\n",
|
|
"print(score_df)\n",
|
|
"ax = score_df.plot.bar(title='statistics_feature')\n",
|
|
"fig = ax.get_figure()\n",
|
|
"#fig.savefig('../figure/base_feature.svg')\n",
|
|
"#print(score_df)"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {
|
|
"collapsed": true
|
|
},
|
|
"outputs": [],
|
|
"source": []
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.6.2"
|
|
}
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 2
|
|
}
|