使用python語言 學習k近鄰分類器的api
歡迎來到我的git查看源代碼: https://github.com/linyi0604/MachineLearning
1 from sklearn.datasets import load_iris 2 from sklearn.cross_validation import train_test_split 3 from sklearn.preprocessing import StandardScaler 4 from sklearn.neighbors import KNeighborsClassifier 5 from sklearn.metrics import classification_report 6 7 ''' 8 k近鄰分類器 9 通過數據的分布對預測數據做出決策 10 屬於無參數估計的一種 11 非常高的計算復雜度和內存消耗 12 ''' 13 14 ''' 15 1 准備數據 16 ''' 17 # 讀取鳶尾花數據集 18 iris = load_iris() 19 # 檢查數據規模 20 # print(iris.data.shape) # (150, 4) 21 # 查看數據說明 22 # print(iris.DESCR) 23 ''' 24 Iris Plants Database 25 ==================== 26 27 Notes 28 ----- 29 Data Set Characteristics: 30 :Number of Instances: 150 (50 in each of three classes) 31 :Number of Attributes: 4 numeric, predictive attributes and the class 32 :Attribute Information: 33 - sepal length in cm 34 - sepal width in cm 35 - petal length in cm 36 - petal width in cm 37 - class: 38 - Iris-Setosa 39 - Iris-Versicolour 40 - Iris-Virginica 41 :Summary Statistics: 42 43 ============== ==== ==== ======= ===== ==================== 44 Min Max Mean SD Class Correlation 45 ============== ==== ==== ======= ===== ==================== 46 sepal length: 4.3 7.9 5.84 0.83 0.7826 47 sepal width: 2.0 4.4 3.05 0.43 -0.4194 48 petal length: 1.0 6.9 3.76 1.76 0.9490 (high!) 49 petal width: 0.1 2.5 1.20 0.76 0.9565 (high!) 50 ============== ==== ==== ======= ===== ==================== 51 52 :Missing Attribute Values: None 53 :Class Distribution: 33.3% for each of 3 classes. 54 :Creator: R.A. Fisher 55 :Donor: Michael Marshall (MARSHALL%PLU@io.arc.nasa.gov) 56 :Date: July, 1988 57 58 This is a copy of UCI ML iris datasets. 59 http://archive.ics.uci.edu/ml/datasets/Iris 60 61 The famous Iris database, first used by Sir R.A Fisher 62 63 This is perhaps the best known database to be found in the 64 pattern recognition literature. Fisher's paper is a classic in the field and 65 is referenced frequently to this day. (See Duda & Hart, for example.) The 66 data set contains 3 classes of 50 instances each, where each class refers to a 67 type of iris plant. One class is linearly separable from the other 2; the 68 latter are NOT linearly separable from each other. 69 70 References 71 ---------- 72 - Fisher,R.A. "The use of multiple measurements in taxonomic problems" 73 Annual Eugenics, 7, Part II, 179-188 (1936); also in "Contributions to 74 Mathematical Statistics" (John Wiley, NY, 1950). 75 - Duda,R.O., & Hart,P.E. (1973) Pattern Classification and Scene Analysis. 76 (Q327.D83) John Wiley & Sons. ISBN 0-471-22361-1. See page 218. 77 - Dasarathy, B.V. (1980) "Nosing Around the Neighborhood: A New System 78 Structure and Classification Rule for Recognition in Partially Exposed 79 Environments". IEEE Transactions on Pattern Analysis and Machine 80 Intelligence, Vol. PAMI-2, No. 1, 67-71. 81 - Gates, G.W. (1972) "The Reduced Nearest Neighbor Rule". IEEE Transactions 82 on Information Theory, May 1972, 431-433. 83 - See also: 1988 MLC Proceedings, 54-64. Cheeseman et al"s AUTOCLASS II 84 conceptual clustering system finds 3 classes in the data. 85 - Many, many more ... 86 87 共有150個數據樣本 88 均勻分布在3個亞種上 89 每個樣本采樣4個花瓣、花萼的形狀描述 90 ''' 91 92 ''' 93 2 划分訓練集合和測試集合 94 ''' 95 x_train, x_test, y_train, y_test = train_test_split(iris.data, 96 iris.target, 97 test_size=0.25, 98 random_state=33) 99 100 ''' 101 3 k近鄰分類器 學習模型和預測 102 ''' 103 # 訓練數據和測試數據進行標准化 104 ss = StandardScaler() 105 x_train = ss.fit_transform(x_train) 106 x_test = ss.transform(x_test) 107 108 # 建立一個k近鄰模型對象 109 knc = KNeighborsClassifier() 110 # 輸入訓練數據進行學習建模 111 knc.fit(x_train, y_train) 112 # 對測試數據進行預測 113 y_predict = knc.predict(x_test) 114 115 ''' 116 4 模型評估 117 ''' 118 print("准確率：", knc.score(x_test, y_test)) 119 print("其他指標:\n", classification_report(y_test, y_predict, target_names=iris.target_names)) 120 ''' 121 准確率： 0.8947368421052632 122 其他指標: 123 precision recall f1-score support 124 125 setosa 1.00 1.00 1.00 8 126 versicolor 0.73 1.00 0.85 11 127 virginica 1.00 0.79 0.88 19 128 129 avg / total 0.92 0.89 0.90 38 130 '''