由於語法渲染問題而影響閱讀體驗， 請移步博客閱讀～
本文GitPage地址

Python Machine Learning

Codes
PDF (English)

  1. sudo pip3.7 install -i https://pypi.tuna.tsinghua.edu.cn/simple sklearn
  2. sudo pip3.7 install -i https://pypi.tuna.tsinghua.edu.cn/simple xgboost
  1. from sklearn.datasets import load_iris
  2. import xgboost as xgb
  3. from xgboost import plot_importance
  4. from matplotlib import pyplot as plt
  5. from sklearn.model_selection import train_test_split
  7. ## read in the iris data
  8. iris = load_iris()
  10. X = iris.data
  11. y = iris.target
  13. X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1234565)
  15. params = {
  16.     'booster': 'gbtree',
  17.     'objective': 'multi:softmax',
  18.     'num_class': 3,
  19.     'gamma': 0.1,
  20.     'max_depth': 6,
  21.     'lambda': 2,
  22.     'subsample': 0.7,
  23.     'colsample_bytree': 0.7,
  24.     'min_child_weight': 3,
  25.     'silent': 1,
  26.     'eta': 0.1,
  27.     'seed': 1000,
  28.     'nthread': 4,
  29. }
  31. plst = params.items()
  34. dtrain = xgb.DMatrix(X_train, y_train)
  35. num_rounds = 500
  36. model = xgb.train(plst, dtrain, num_rounds)
  38. ## 对测试集进行预测
  39. dtest = xgb.DMatrix(X_test)
  40. ans = model.predict(dtest)
  42. ## 计算准确率
  43. cnt1 = 0
  44. cnt2 = 0
  45. for i in range(len(y_test)):
  46.     if ans[i] == y_test[i]:
  47.         cnt1 += 1
  48.     else:
  49.         cnt2 += 1
  51. print("Accuracy: %.2f %% " % (100 * cnt1 / (cnt1 + cnt2)))
  53. ## 显示重要特征
  54. plot_importance(model)
  55. plt.show()

Enjoy~

本文由Python腳本GitHub/語雀自動更新

由於語法渲染問題而影響閱讀體驗， 請移步博客閱讀～
本文GitPage地址

GitHub: Karobben
Blog:Karobben
BiliBili:史上最不正經的生物狗