使用 KNN 的 IBM HR Analytics 员工流失率和绩效

人员流失是一个影响所有企业的问题，无论其地理位置、行业和公司规模如何。这是一个组织的一个主要问题，在许多组织中，预测离职率是人力资源 (HR) 需求的最前沿。组织面临着因员工流失而产生的巨额成本。随着机器学习和数据科学的进步，可以预测员工流失率，我们将使用 KNN（k-nearest neighbours）算法进行预测。
数据集：
IBM 人力资源部门发布的数据集可在 Kaggle 上获得。
数据集
代码：用于分类的 KNN 算法的实现。
加载库

Python3

# performing linear algebra
import numpy as np
 
# data processing
import pandas as pd
 
# visualisation
import matplotlib.pyplot as plt
import seaborn as sns % matplotlib inline

代码：导入数据集

Python3

dataset = pd.read_csv("WA_Fn-UseC_-HR-Employee-Attrition.csv")
print (dataset.head)

输出 ：

代码：有关数据集的信息

Python3

df.info()

输出 ：

RangeIndex: 1470 entries, 0 to 1469
Data columns (total 35 columns):
Age                         1470 non-null int64
Attrition                   1470 non-null object
BusinessTravel              1470 non-null object
DailyRate                   1470 non-null int64
Department                  1470 non-null object
DistanceFromHome            1470 non-null int64
Education                   1470 non-null int64
EducationField              1470 non-null object
EmployeeCount               1470 non-null int64
EmployeeNumber              1470 non-null int64
EnvironmentSatisfaction     1470 non-null int64
Gender                      1470 non-null object
HourlyRate                  1470 non-null int64
JobInvolvement              1470 non-null int64
JobLevel                    1470 non-null int64
JobRole                     1470 non-null object
JobSatisfaction             1470 non-null int64
MaritalStatus               1470 non-null object
MonthlyIncome               1470 non-null int64
MonthlyRate                 1470 non-null int64
NumCompaniesWorked          1470 non-null int64
Over18                      1470 non-null object
OverTime                    1470 non-null object
PercentSalaryHike           1470 non-null int64
PerformanceRating           1470 non-null int64
RelationshipSatisfaction    1470 non-null int64
StandardHours               1470 non-null int64
StockOptionLevel            1470 non-null int64
TotalWorkingYears           1470 non-null int64
TrainingTimesLastYear       1470 non-null int64
WorkLifeBalance             1470 non-null int64
YearsAtCompany              1470 non-null int64
YearsInCurrentRole          1470 non-null int64
YearsSinceLastPromotion     1470 non-null int64
YearsWithCurrManager        1470 non-null int64
dtypes: int64(26), object(9)
memory usage: 402.0+ KB

代码：可视化数据

Python3

# heatmap to check the missing value
plt.figure(figsize =(10, 4))
sns.heatmap(dataset.isnull(), yticklabels = False, cbar = False, cmap ='viridis')

输出：

因此，我们可以看到数据集中没有缺失值。
这是一个二元分类问题，因此 2 个类之间的实例分布如下所示：

Python3

sns.set_style('darkgrid')
sns.countplot(x ='Attrition', data = dataset)

输出：

代码：

Python3

sns.lmplot(x = 'Age', y = 'DailyRate', hue = 'Attrition', data = dataset)

输出：

代码 ：

Python3

plt.figure(figsize =(10, 6))
sns.boxplot(y ='MonthlyIncome', x ='Attrition', data = dataset)

输出：

预处理数据
数据集中有 4 个不相关的列，即：EmployeeCount、EmployeeNumber、Over18 和 StandardHour。因此，我们必须删除这些以获得更高的准确性。
代码：

Python3

dataset.drop('EmployeeCount', axis = 1, inplace = True)
dataset.drop('StandardHours', axis = 1, inplace = True)
dataset.drop('EmployeeNumber', axis = 1, inplace = True)
dataset.drop('Over18', axis = 1, inplace = True)
print(dataset.shape)

输出：

(1470, 31)

因此，我们删除了不相关的列。
代码：输入和输出数据

Python3

y = dataset.iloc[:, 1]
X = dataset
X.drop('Attrition', axis = 1, inplace = True)

代码：标签编码

Python3

from sklearn.preprocessing import LabelEncoder
lb = LabelEncoder()
y = lb.fit_transform(y)

在数据集中有 7 个分类数据，所以我们必须将它们更改为 int 数据，即我们必须创建 7 个虚拟变量以获得更好的准确性。
代码：虚拟变量创建

Python3

dum_BusinessTravel = pd.get_dummies(dataset['BusinessTravel'],
                                    prefix ='BusinessTravel')
dum_Department = pd.get_dummies(dataset['Department'],
                                prefix ='Department')
dum_EducationField = pd.get_dummies(dataset['EducationField'],
                                    prefix ='EducationField')
dum_Gender = pd.get_dummies(dataset['Gender'],
                            prefix ='Gender', drop_first = True)
dum_JobRole = pd.get_dummies(dataset['JobRole'],
                             prefix ='JobRole')
dum_MaritalStatus = pd.get_dummies(dataset['MaritalStatus'],
                                   prefix ='MaritalStatus')
dum_OverTime = pd.get_dummies(dataset['OverTime'],
                              prefix ='OverTime', drop_first = True)
# Adding these dummy variable to input X
X = pd.concat([x, dum_BusinessTravel, dum_Department,
               dum_EducationField, dum_Gender, dum_JobRole,
               dum_MaritalStatus, dum_OverTime], axis = 1)
# Removing the categorical data
X.drop(['BusinessTravel', 'Department', 'EducationField',
        'Gender', 'JobRole', 'MaritalStatus', 'OverTime'],
        axis = 1, inplace = True)
 
print(X.shape)
print(y.shape)

输出：

(1470, 49)
(1470, )

代码：将数据拆分为训练和测试

Python3

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size = 0.25, random_state = 40)

至此，预处理完成，现在我们必须将 KNN 应用于数据集。
模型执行代码：使用 KNeighborsClassifier 在错误分类错误的帮助下找到最佳邻居数。

Python3

from sklearn.neighbors import KNeighborsClassifier
neighbors = []
cv_scores = []
   
from sklearn.model_selection import cross_val_score
# perform 10 fold cross validation
for k in range(1, 40, 2):
    neighbors.append(k)
    knn = KNeighborsClassifier(n_neighbors = k)
    scores = cross_val_score(
        knn, X_train, y_train, cv = 10, scoring = 'accuracy')
    cv_scores.append(scores.mean())
error_rate = [1-x for x in cv_scores]
   
# determining the best k
optimal_k = neighbors[error_rate.index(min(error_rate))]
print('The optimal number of neighbors is % d ' % optimal_k)
   
# plot misclassification error versus k
plt.figure(figsize = (10, 6))
plt.plot(range(1, 40, 2), error_rate, color ='blue', linestyle ='dashed', marker ='o',
         markerfacecolor ='red', markersize = 10)
plt.xlabel('Number of neighbors')
plt.ylabel('Misclassification Error')
plt.show()

输出：

The optimal number of neighbors is  7  

代码：预测分数

Python3

from sklearn.model_selection import cross_val_predict, cross_val_score
from sklearn.metrics import accuracy_score, classification_report
from sklearn.metrics import confusion_matrix
 
def print_score(clf, X_train, y_train, X_test, y_test, train = True):
    if train:
        print("Train Result:")
        print("------------")
        print("Classification Report: \n {}\n".format(classification_report(
                y_train, clf.predict(X_train))))
        print("Confusion Matrix: \n {}\n".format(confusion_matrix(
                y_train, clf.predict(X_train))))
 
        res = cross_val_score(clf, X_train, y_train,
                              cv = 10, scoring ='accuracy')
        print("Average Accuracy: \t {0:.4f}".format(np.mean(res)))
        print("Accuracy SD: \t\t {0:.4f}".format(np.std(res)))
        print("accuracy score: {0:.4f}\n".format(accuracy_score(
                y_train, clf.predict(X_train))))
        print("----------------------------------------------------------")
                
    elif train == False:
        print("Test Result:")
        print("-----------")
        print("Classification Report: \n {}\n".format(
                classification_report(y_test, clf.predict(X_test))))
        print("Confusion Matrix: \n {}\n".format(
                confusion_matrix(y_test, clf.predict(X_test))))
        print("accuracy score: {0:.4f}\n".format(
                accuracy_score(y_test, clf.predict(X_test))))
        print("-----------------------------------------------------------")
         
knn = KNeighborsClassifier(n_neighbors = 7)
knn.fit(X_train, y_train)
print_score(knn, X_train, y_train, X_test, y_test, train = True)
print_score(knn, X_train, y_train, X_test, y_test, train = False)

输出：

Train Result:
------------
Classification Report: 
               precision    recall  f1-score   support

           0       0.86      0.99      0.92       922
           1       0.83      0.19      0.32       180

    accuracy                           0.86      1102
   macro avg       0.85      0.59      0.62      1102
weighted avg       0.86      0.86      0.82      1102


Confusion Matrix: 
 [[915   7]
 [145  35]]

Average Accuracy:      0.8421
Accuracy SD:          0.0148
accuracy score: 0.8621

-----------------------------------------------------------
Test Result:
-----------
Classification Report: 
               precision    recall  f1-score   support

           0       0.84      0.96      0.90       311
           1       0.14      0.04      0.06        57

    accuracy                           0.82       368
   macro avg       0.49      0.50      0.48       368
weighted avg       0.74      0.82      0.77       368


Confusion Matrix: 
 [[299  12]
 [ 55   2]]

accuracy score: 0.8179