sklearn神经网络/BP神经网络实现葡萄酒分类问题

痛定思痛。 2023-07-18 15:28 43阅读 0赞

葡萄酒分类问题：数据集中葡萄酒三个种类，13种特征

1. Title of Database: Wine recognition data
    	Updated Sept 21, 1998 by C.Blake : Added attribute information
    
    2. Sources:
       (a) Forina, M. et al, PARVUS - An Extendible Package for Data
           Exploration, Classification and Correlation. Institute of Pharmaceutical
           and Food Analysis and Technologies, Via Brigata Salerno, 
           16147 Genoa, Italy.
    
       (b) Stefan Aeberhard, email: stefan@coral.cs.jcu.edu.au
       (c) July 1991
    3. Past Usage:
    
       (1)
       S. Aeberhard, D. Coomans and O. de Vel,
       Comparison of Classifiers in High Dimensional Settings,
       Tech. Rep. no. 92-02, (1992), Dept. of Computer Science and Dept. of
       Mathematics and Statistics, James Cook University of North Queensland.
       (Also submitted to Technometrics).
    
       The data was used with many others for comparing various 
       classifiers. The classes are separable, though only RDA 
       has achieved 100% correct classification.
       (RDA : 100%, QDA 99.4%, LDA 98.9%, 1NN 96.1% (z-transformed data))
       (All results using the leave-one-out technique)
    
       In a classification context, this is a well posed problem 
       with "well behaved" class structures. A good data set 
       for first testing of a new classifier, but not very 
       challenging.
    
       (2) 
       S. Aeberhard, D. Coomans and O. de Vel,
       "THE CLASSIFICATION PERFORMANCE OF RDA"
       Tech. Rep. no. 92-01, (1992), Dept. of Computer Science and Dept. of
       Mathematics and Statistics, James Cook University of North Queensland.
       (Also submitted to Journal of Chemometrics).
    
       Here, the data was used to illustrate the superior performance of
       the use of a new appreciation function with RDA. 
    
    4. Relevant Information:
    
       -- These data are the results of a chemical analysis of
          wines grown in the same region in Italy but derived from three
          different cultivars.
          The analysis determined the quantities of 13 constituents
          found in each of the three types of wines. 
    
       -- I think that the initial data set had around 30 variables, but 
          for some reason I only have the 13 dimensional version. 
          I had a list of what the 30 or so variables were, but a.) 
          I lost it, and b.), I would not know which 13 variables
          are included in the set.
    
       -- The attributes are (dontated by Riccardo Leardi, 
    	riclea@anchem.unige.it )
     	1) Alcohol
     	2) Malic acid
     	3) Ash
    	4) Alcalinity of ash  
     	5) Magnesium
    	6) Total phenols
     	7) Flavanoids
     	8) Nonflavanoid phenols
     	9) Proanthocyanins
    	10)Color intensity
     	11)Hue
     	12)OD280/OD315 of diluted wines
     	13)Proline            
    
    5. Number of Instances
    
          	class 1 59
    	class 2 71
    	class 3 48
    
    6. Number of Attributes 
    	
    	13
    
    7. For Each Attribute:
    
    	All attributes are continuous
    	
    	No statistics available, but suggest to standardise
    	variables for certain uses (e.g. for us with classifiers
    	which are NOT scale invariant)
    
    	NOTE: 1st attribute is class identifier (1-3)
    
    8. Missing Attribute Values:
    
    	None
    
    9. Class Distribution: number of instances per class
    
          	class 1 59
    	class 2 71
    	class 3 48

相关csv数据下载：  
[https://download.csdn.net/download/ayangann915/12278359][https_download.csdn.net_download_ayangann915_12278359]

import numpy as np
    from sklearn.model_selection import train_test_split
    from sklearn.preprocessing import StandardScaler
    from sklearn.neural_network import MLPClassifier
    from sklearn.metrics import classification_report,confusion_matrix
    
    #读数据
    wine = np.genfromtxt("wine_data.csv", delimiter=",")
    #取所有的行，和第一列之后的数据，因为第一列是标签，后面的是特征
    X = wine[:,1:]
    #标签第一列
    y = wine[:,0]
    
    #划分训练集和测试集，默认四分之一测试集，四分之三训练集
    x_train, x_test, y_train, y_test = train_test_split(X, y, test_size=0.3)
    #标准化数据
    scaler = StandardScaler()
    x_train = scaler.fit_transform(x_train)
    x_test = scaler.fit_transform(x_test)
    #构建神经网络，其中三个隐藏层，分别有100，50，20个神经元，最大训练次数500
    mlp = MLPClassifier(hidden_layer_sizes=(100,50,20),max_iter=500)
    #训练
    mlp.fit(x_train, y_train)
    #预测
    predict = mlp.predict(x_test)
    #打印测试结果和真实标签的准确率
    print(classification_report(predict, y_test))
    
    
    """ precision recall f1-score support 1.0 0.94 1.00 0.97 16 2.0 1.00 0.95 0.97 19 3.0 1.00 1.00 1.00 19 accuracy 0.98 54 macro avg 0.98 0.98 0.98 54 weighted avg 0.98 0.98 0.98 54 """

搭建BP神经网络（准确率较低）

from sklearn.model_selection import train_test_split
    from sklearn.preprocessing import LabelBinarizer, StandardScaler
    from sklearn.metrics import confusion_matrix, classification_report
    import numpy as np
    
    wine = np.genfromtxt("wine_data.csv",delimiter=",")
    X = wine[:, 1:]
    y = wine[:, 0]
    x_train, x_test, y_train, y_test = train_test_split(X, y)
    # scale = StandardScaler()
    # x_train = scale.fit_transform(x_train)
    # x_test = scale.fit_transform(x_test)
    rate = 0.11
    #权值，13输入，3输出中间隐藏层100个神经元
    v = np.random.random((13,100))*2-1
    w = np.random.random((100,3))*2-1
    
    label_train = LabelBinarizer().fit_transform(y_train)
    
    #定义激活函数及其导数
    def sigmoid(x):
        return 1/(1+np.exp(-x))
    def dsigmoid(x):
        return x*(1-x)
    
    #训练
    def train(X, y, step = 10000,rate=0.11):
        global v,w
        for n in range(step+1):
            i = np.random.randint(X.shape[0])
            x = X[i]
            x = np.atleast_2d(x)
            
            L1 = sigmoid(np.dot(x,v))
            L2 = sigmoid(np.dot(L1,w))
            
            L2_delta = (y[i]-L2)*dsigmoid(L2)
            L1_delta = L2_delta.dot(w.T)*dsigmoid(L1)
            
            w_ = rate * L1.T.dot(L2_delta)
            v_ = rate * x.T.dot(L1_delta)
            w = w + w_
            v = v + v_
    
            if n%1000 == 0 :
                output = predict(x_test)
                predictions = np.argmax(output, axis=1)
                acc = np.mean(np.equal(predictions, y_test))
                print("step:",n,"acc:",acc)
                
                
    def predict(x):
        L1 = sigmoid(np.dot(x,v))
        L2 = sigmoid(np.dot(L1,w))
        return L2
    
    train(x_train,label_train,30000)

[https_download.csdn.net_download_ayangann915_12278359]: https://download.csdn.net/download/ayangann915/12278359