Question

我正在使用svm来查看我是否可以获取棒球数据并分类击中次数并估算本垒打。当我多次运行模型时，似乎得到了不同的结果，结果，我进行了一次仿真，并运行了100次模型，但是我不明白为什么以及是什么导致了变化。有人可以解释为什么会这样吗？我确实设置了random_state = 42

import pandas as pd
from mlxtend.plotting import plot_decision_regions
import matplotlib.pyplot as plt
from sklearn.svm import SVC
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score, f1_score, precision_score, recall_score, classification_report, confusion_matrix
from sklearn import metrics
import statistics
import numpy as np
result_array = []
players = [488768,  517369, 461314, 477165, 506560, 572114, 641319, 592669, 622534, 605486, 602922, 518466, 572362, 519082, 623182, 595978, 543272]

dfSave = pd.DataFrame(columns=['Mean','Max','Min','Std', 'Accuracy', 'Precision', 'f1_score', 'Recall_Score', 'First_Name', 'Last_Name'])
for i in players:
    batter = i

    df = pd.read_csv('D:baseballData_2016_use.csv')

    df2 = pd.read_csv('D:padres_2016_home.csv')  #Team to test

    dataFilter = df.loc[df['Home_Team'] == 'Orioles'] #Stadium to train model to.

    dataFilter2 = df2.loc[df2['Batter_ID'] == batter] #Players to test in stadium

    j = 0
    while j <= 100:



     predict = dataFilter2.iloc[:,[4,5]]


     X =dataFilter.iloc[:,[4,5]]
     y = dataFilter.iloc[:,3]
     y = y.astype(np.integer)

     X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.30)



     svclassifier = SVC(C=4, cache_size=200, class_weight= None, coef0=0.0,
    decision_function_shape='ovo', degree=3, gamma=0.001, kernel='rbf',
    max_iter=-1, probability=False, random_state=42, shrinking=False,
    tol=0.001, verbose=False) #defaults

     svclassifier.fit(X_train, y_train)

     y_pred = svclassifier.predict(X_test)



     predicted= svclassifier.predict(predict)


     listDf = []

     sum = 0  # print predicted home runs
     for i in predicted:
      if i == 1:
       sum = sum + 1
     result_array.append(sum)
     print(sum)


     j = j + 1

    firstName = dataFilter2.loc[1:,'Batter_First_Name'].values
    lastName = dataFilter2.loc[1:,'Batter_Last_Name'].values
    listDf.append({'Mean': statistics.mean(result_array),'Max' : max(result_array),'Min' : min(result_array),'Std' : statistics.stdev(result_array),
                   'Accuracy' : metrics.accuracy_score(y_test, y_pred), 'Precision' : precision_score(y_test, y_pred, average="macro"), 'f1_score' :f1_score(y_test, y_pred, average="macro"),
                   "Recall_Score" : recall_score(y_test, y_pred, average="macro"), 'First_Name' :  firstName[0],'Last_Name' :  lastName[0]})
    dfSave = pd.DataFrame(listDf)


    dfSave.to_csv('D:test9999.csv', mode='a')
    result_array = []

Answer 1

在您的代码中，随机性来自train_test_split，每次运行均给出不同的划分。

您可以通过修复random_state来避免这种情况，但是，最好多次运行（如您所愿），获取输出得分的分布calculate the confidence interval on the score并报告，这是一种更好的做法。 / p>

支持向量机随机

1 个答案: