5 Employee_Appraisal_Regression_Analysis.ipynb

Kernel: Python 3 (ipykernel)

Employee Appraisal Regression Analysis

This notebook explores a dataset with employee appraisal data and builds a Linear Regression model.

Features used:

• Rating

• Behavior

• Experience

• Band (encoded as Band_Num)

Target:

• Appraisal

# 1. Import Libraries
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.preprocessing import LabelEncoder
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split, cross_val_score, KFold
from sklearn.metrics import mean_squared_error, r2_score

# 2. Load the Data
df = pd.read_csv('appraisal_dataset.csv')
df.head()

Dataset Overview

df.shape 

(1023, 6)

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 1023 entries, 0 to 1022
Data columns (total 6 columns):
 #   Column      Non-Null Count  Dtype  
---  ------      --------------  -----  
 0   Rating      1023 non-null   float64
 1   Behavior    1023 non-null   float64
 2   Experience  1023 non-null   int64  
 3   Band        1023 non-null   object 
 4   Band_Num    1023 non-null   int64  
 5   Appraisal   1023 non-null   float64
dtypes: float64(3), int64(2), object(1)
memory usage: 48.1+ KB

df.describe()

df.describe(include="object")

#df.isnull().sum()

Encode Categorical Feature (Band)

from sklearn.preprocessing import LabelEncoder
le = LabelEncoder()
df['Band_Num'] = le.fit_transform(df['Band'])
df[['Band', 'Band_Num']].drop_duplicates()

Visuals

plt.figure(figsize=(11, 8))
sns.heatmap(df.corr(numeric_only=True), annot=True, cmap='coolwarm')
plt.title("Correlation Matrix (Numerical Columns Only)")
plt.show()

sns.boxplot(x='Band', y='Appraisal', data=df)
plt.title("Appraisal by Band")
plt.show()

Data Preparation

X = df[['Rating', 'Behavior', 'Experience', 'Band_Num']]
y = df['Appraisal']

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

model trained( X_train and y_train)
ypred=model(xtest)
ytest - ypred 

Train Linear Regression Model

Appraisal_pred = LinearRegression()
Appraisal_pred.fit(X_train, y_train)
y_pred = Appraisal_pred.predict(X_test)

print("R2 Score:", r2_score(y_test, y_pred))
print("MSE:", mean_squared_error(y_test, y_pred))
print("Coefficients:", Appraisal_pred.coef_)
print("Intercept:", Appraisal_pred.intercept_)

R2 Score: 0.8713438627272418
MSE: 24.88148233009516
Coefficients: [10.18785705  4.8082995   1.51022801  2.92957254]
Intercept: 0.08144186829784417

Linear equation for Salary Prediction

Appraisal = 10.2(rating) + 5(Exp) +1.5(be)+3(B_num) + 0.081

K-Fold Cross-Validation

K-Fold Cross Validation is a technique used to evaluate the performance of a machine learning model by splitting the dataset into K equal parts (folds):

• The model is trained on K-1 folds and validated on the remaining fold.

• This process is repeated K times, each time using a different fold as the validation set.

• The final performance is the average of all K validation scores.

Why use it?

• It helps reduce overfitting.

• Provides a more robust estimate of model performance.

• Ensures that every data point is used for both training and validation.

kf = KFold(n_splits=5, shuffle=True, random_state=42)
scores = cross_val_score(Appraisal_pred , X, y, cv=kf, scoring='r2')

print("K-Fold R2 Scores:", scores)
print("Average R2 Score:", np.mean(scores))

K-Fold R2 Scores: [0.87134386 0.87550696 0.88419112 0.88011813 0.86527841]
Average R2 Score: 0.8752876954373467

# Visualizing R2 score variation across folds
plt.figure(figsize=(8, 5))
plt.plot(range(1, len(scores)+1), scores, marker='o', linestyle='--', color='blue')
plt.title("R² Score per Fold")
plt.xlabel("Fold")
plt.ylabel("R² Score")
plt.ylim(0, 1)
plt.grid(True)
plt.xticks(range(1, len(scores)+1))
plt.show()

# Alternatively, use barplot
plt.figure(figsize=(8, 5))
sns.barplot(x=list(range(1, len(scores)+1)), y=scores, palette='viridis')
plt.title("R² Score Variation Across Folds")
plt.xlabel("Fold")
plt.ylabel("R² Score")
plt.ylim(0, 1)
plt.show()

Insights & Conclusion

• Experience and Rating positively impact Appraisal.

• Appraisal is somewhat influenced by Band.

• Linear Regression performs reasonably well.

• K-Fold helps validate the model across different subsets.

Quick Practice:

• Check Model Performance taling two paramters, Three parameters,

• Vary test and train Split

• Name your Model as Appraisalsys

Prediction to New Data

# Step 1: Load new data
new_df = pd.read_csv('NewData.csv')
new_df.head()

# Step 2: Encode the 'Band' column using the same LabelEncoder
# Important: Use the SAME label encoder instance used in training
# If not available, you can refit with the same category order

# Example: manually set the encoder based on original training data
band_classes = ['A', 'B', 'C', 'D']  # Replace with actual categories from training
le = LabelEncoder()
le.fit(band_classes)

new_df['Band_Num'] = le.transform(new_df['Band'])

# Step 3: Select relevant features
X_new = new_df[['Rating', 'Behavior', 'Experience', 'Band_Num']]

# Step 4: Predict using the trained model
new_predictions = Appraisal_pred .predict(X_new)
new_df['Predicted_Appraisal'] = new_predictions

# View results
new_df[['Rating', 'Behavior', 'Experience', 'Band', 'Predicted_Appraisal']]

new_df.to_csv('predicted_appraisals.csv', index=False)
print("Predictions saved to predicted_appraisals.csv")

Predictions saved to predicted_appraisals.csv