diff --git a/train.py b/train.py
new file mode 100644
index 0000000..fefe958
--- /dev/null
+++ b/train.py
@@ -0,0 +1,129 @@
+from sklearn.svm import SVC
+from sklearn.neighbors import KNeighborsClassifier  # K近邻分类器
+from sklearn.tree import DecisionTreeClassifier  # 决策树分类器
+from sklearn.naive_bayes import GaussianNB  # 高斯朴素贝叶斯函数
+from sklearn.neural_network import MLPClassifier
+import numpy as np
+
+from sklearn.decomposition import PCA
+from sklearn.model_selection import GridSearchCV # 调节参数
+import random
+import time
+import csv
+
+import logging
+import matplotlib.pyplot as plt
+
+logging.basicConfig(level=logging.INFO)
+
+
+# 平均值
+def getXmean(x_train:list):
+    return np.mean(x_train, axis=0)
+
+# 归一化
+def centralized(x_test:list, mean:list):
+    x_test -= mean
+    return x_test
+
+def preData(x_train:list, y_train:list, x_test:list, y_test:list):
+
+    if len(x_train) != len(y_train) or len(x_test) != len(y_test):
+        return
+
+    if len(x_train) == 0 or len(x_test) == 0:
+        return
+
+    # 归一化
+    # 暂时不需要，以后看情况
+    # mean = getXmean(x_train)
+    # x_train = centralized(x_train, mean)
+    # x_test = centralized(x_test, mean)
+
+    cc = list(zip(x_train, y_train))
+    random.shuffle(cc)
+    x_train[:], y_train[:] = zip(*cc)
+
+    cc = list(zip(x_test, y_test))
+    random.shuffle(cc)
+    x_test[:], y_test[:] = zip(*cc)
+
+    # 降维
+    if len(x_train + y_train) < 150:
+        n = len(x_train + y_train)
+    else:
+        n = 150
+        
+    pca = PCA(n_components=n, # 主成分
+              svd_solver='randomized', # 随机打乱
+              whiten=True) # 白化
+    # print(type(x_train))
+    tmp = np.concatenate((x_train, x_test), axis=0)
+    pca.fit(tmp)
+
+    x_train = pca.transform(x_train)
+    x_test = pca.transform(x_test)
+    
+    return x_train, y_train, x_test, y_test
+
+# 准确率
+def accuracy(y_true:list, y_pred:list):
+    if len(y_true) != len(y_pred):
+        return
+    num = 0
+    for i in range(len(y_pred)):
+        if y_pred[i] == y_true[i]:
+            num += 1
+    return num/len(y_pred)
+
+
+# 如果每个节点都用一种模型，则每个节点训练的效果都会基本相同，
+# 所以给出几种不同的模型
+# 返回模型、准确率和训练时间
+# 准确率相同的情况下，训练时间越少效果越好
+
+def svcTrain(x_train:list, y_train:list, x_test:list, y_test:list):
+    x_train, y_train, x_test, y_test = preData(x_train, y_train, x_test, y_test)
+
+    start = time.time()
+    start = int(round(start * 1000))
+
+    svc_last = SVC(C=5.0, gamma=0.001, max_iter=-1)
+    svc = svc_last.fit(x_train, y_train)
+
+    end = time.time()
+    end = int(round(end * 1000))
+
+    answer_svc = svc_last.predict(x_test)
+    return svc, accuracy(y_test, answer_svc), end-start
+
+
+def gnbTrain(x_train:list, y_train:list, x_test:list, y_test:list):
+    x_train, y_train, x_test, y_test = preData(x_train, y_train, x_test, y_test)
+    
+    start = time.time()
+    start = int(round(start * 1000))
+
+    gnb = GaussianNB().fit(x_train, y_train)
+
+    end = time.time()
+    end = int(round(end * 1000))
+
+    answer_gnb = gnb.predict(x_test)
+    return gnb, accuracy(y_test, answer_gnb), end-start
+
+
+def mlpTrain(x_train:list, y_train:list, x_test:list, y_test:list):
+    # 适用于低维
+    x_train, y_train, x_test, y_test = preData(x_train, y_train, x_test, y_test)
+    
+    start = time.time()
+    start = int(round(start * 1000))
+
+    mlp = MLPClassifier().fit(x_train, y_train)
+    
+    end = time.time()
+    end = int(round(end * 1000))
+
+    answer_mlp = mlp.predict(x_test)
+    return mlp, accuracy(y_test, answer_mlp), end-start