College Student Placement Prediction

A machine learning project that predicts whether a college student will get placed based on their CGPA, IQ, and internship experience using Logistic Regression.

Dataset

The dataset (college_student_placement_dataset.csv) contains the following features:

• College_ID: Unique identifier for each student
• IQ: Intelligence Quotient score
• CGPA: Cumulative Grade Point Average
• Internship_Experience: Whether the student has internship experience (Yes/No)
• Placement: Target variable - whether the student got placed (Yes/No)

Sample Data:

College_ID	IQ	CGPA	Internship_Experience	Placement
CLG0030	107	6.28	No	No
CLG0061	97	5.37	No	No
CLG0036	109	5.83	No	No
CLG0055	122	5.75	Yes	No

Steps Performed

1. Data Loading & Exploration

import pandas as pd
import numpy as np

df = pd.read_csv('college_student_placement_dataset.csv')
df.head()

Output: First 5 rows of the dataset displayed.

2. Data Preprocessing

# Convert categorical variables to numerical
df["Internship_Experience"] = df["Internship_Experience"].map({"Yes": 1, "No": 0})
df["Placement"] = df["Placement"].map({"Yes": 1, "No": 0})

# Sort by College_ID
df = df.sort_values(by="College_ID", ascending=True)

df.head()
len(df)

Output:

• Categorical columns converted to binary (0/1)
• Dataset sorted by College_ID
• Total number of samples displayed

3. Exploratory Data Analysis (Visualization)

import matplotlib.pyplot as plt

plt.scatter(df["CGPA"], df["IQ"], c=df["Placement"])
plt.title("IQ vs CGPA")
plt.xlabel("CGPA")
plt.ylabel("IQ")

Output: Scatter plot showing IQ vs CGPA colored by placement status (0=Not Placed, 1=Placed)

4. Feature Selection

x = df.iloc[:, 1:3]  # Features: IQ, CGPA
y = df.iloc[:, -1]   # Target: Placement

Note: Internship_Experience was excluded from features (only IQ and CGPA used)

5. Train-Test Split

from sklearn.model_selection import train_test_split

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.1)

Output: 90% training data, 10% test data

6. Feature Scaling

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)

Output: Standardized features (mean=0, std=1) for both train and test sets

7. Model Training (Logistic Regression)

from sklearn.linear_model import LogisticRegression

clf = LogisticRegression()
clf.fit(x_train, y_train)

Output: Logistic Regression model trained successfully

8. Model Evaluation

y_pred = clf.predict(x_test)
y_test

from sklearn.metrics import accuracy_score
accuracy_score(y_test, y_pred)

Output: Accuracy score on test set displayed

9. Decision Boundary Visualization

from mlxtend.plotting import plot_decision_regions

plot_decision_regions(x_train, y_train.values, clf=clf, legend=2)

Output: Decision boundary plot showing how the model separates placed vs not placed students based on IQ and CGPA

10. Model Persistence

import pickle

pickle.dump(clf, open('model.pkl', 'wb'))

Output: Trained model saved as model.pkl for future use

How to Run

1. Install dependencies:

pip install pandas numpy matplotlib scikit-learn mlxtend

2. Run the notebook:

jupyter notebook main.ipynb

3. Use the saved model:

import pickle
import numpy as np
from sklearn.preprocessing import StandardScaler

# Load model
model = pickle.load(open('model.pkl', 'rb'))

# Prepare new data (IQ, CGPA)
new_data = np.array([[110, 7.5]])  # Example: IQ=110, CGPA=7.5

# Scale using same scaler (you'll need to fit on training data first)
# For production, save the scaler too!
scaler = StandardScaler()
# ... fit scaler on training data ...
new_data_scaled = scaler.transform(new_data)

# Predict
prediction = model.predict(new_data_scaled)
print("Placed" if prediction[0] == 1 else "Not Placed")

Results Summary

Metric	Value
Algorithm	Logistic Regression
Features Used	IQ, CGPA
Test Size	10%
Preprocessing	StandardScaler
Model Saved	model.pkl

Future Improvements

• Include Internship_Experience as a feature
• Try other algorithms (Random Forest, SVM, XGBoost)
• Perform hyperparameter tuning
• Add cross-validation
• Save the StandardScaler along with the model
• Create a simple web interface (Streamlit/Flask)

License

This project is for educational purposes.