Predictive Modeling for Click-Through Rate Optimization

Internship Minor Project | Logistic Regression Model (using scikit-learn)

---

Overview

This project predicts whether a user will click on an online advertisement using Logistic Regression.
By analyzing user demographics and behavior data, the model helps digital marketers target the right users, improving Click-Through Rate (CTR) and Return on Ad Spend (ROAS).

---

Problem Statement

ConnectSphere Digital runs online campaigns but faces inefficient ad spending — ads are shown to users unlikely to engage.
The aim is to use machine learning to predict users likely to click ads, helping optimize marketing budgets and performance.

---

Dataset

Dataset Sources

• Local File: advertising.csv
• Google Drive: Download Dataset

Dataset Description

Feature	Description
Age	Age of the user
Area Income	Average income of the user's geographical area
Daily Internet Usage	Average daily internet usage (minutes)
Daily Time Spent on Site	Average minutes user spends on the website daily
Clicked on Ad	Target variable (1 = Clicked, 0 = Not Clicked)

---

Implementation

Libraries Used

pandas
scikit-learn