A company wants to invest a large budget in an online advertising campaign; however, the click-through rate (CTR) is still low, at only about 50%. This low CTR indicates that most ad impressions do not capture customers' attention. Therefore, a data-driven approach using machine learning is needed to understand the characteristics of customers who are likely to click on ads. With these insights, the company can develop targeted strategies to significantly improve the CTR.
The absence of a predictive system capable of identifying customers who are more likely to click on ads leads to suboptimal allocation of the advertising budget. This also reduces the effectiveness of campaigns, as ads are often shown to a less relevant audience. Additionally, the lack of insights into the features that influence ad clicks makes it difficult for the company to design more effective marketing strategies.
This project aims to build a machine learning-based classification model that can predict the likelihood of customers clicking on ads based on behavioral and demographic data, such as time spent on the site, daily internet usage, and income. With this model, the company can improve the click-through rate (CTR) through more precise targeting strategies.
- Define the problem, goal, objectives, model, and metrics.
- Import the necessary data and libraries.
- Perform exploratory data analysis (EDA) to find valuable insights.
- Conduct data preprocessing, such as handling null values, outliers, and more.
- Build models using various classification techniques.
- Determine the best model based on evaluation metrics.
- Make recommendations based on the findings from modeling and EDA.
This Dataset has 1000 columns and 11 columns.
| Column Name | Description |
|---|---|
| Unnamed: 0 | ID of Customers |
| Daily Time Spent on Site | Time spent by the user on a site in minutes |
| Age | Customer’s age in terms of years |
| Area Income | Average income of geographical area of consumer |
| Daily Internet Usage | Average minutes in a day consumer is on the internet |
| Male | Gender of the customer |
| Timestamp | Time at which user clicked on an Ad or the closed window |
| Clicked on Ad | Whether or not the customer clicked on an Ad (Target Variable) |
| City | City of the consumer |
| Province | Province of the consumer |
| Category | Category of the advertisement |
- Click Through Rate (CTR)
- Accuracy dan ROC-AUC score.
- Python Programming Language
- Jupyterlab