https://github.com/himalayaashish/fakeadclicksdetection
This project focuses on detecting fake ad clicks using advanced machine learning techniques. I applied models to classify legitimate and fraudulent clicks based on patterns and user behavior data. The system leverages data preprocessing, model training, and evaluation to optimize the detection process for real-time applications.
https://github.com/himalayaashish/fakeadclicksdetection
classification machine-learning tensorflow tensorflow2
Last synced: 2 months ago
JSON representation
This project focuses on detecting fake ad clicks using advanced machine learning techniques. I applied models to classify legitimate and fraudulent clicks based on patterns and user behavior data. The system leverages data preprocessing, model training, and evaluation to optimize the detection process for real-time applications.
- Host: GitHub
- URL: https://github.com/himalayaashish/fakeadclicksdetection
- Owner: himalayaashish
- Created: 2022-05-12T23:54:19.000Z (about 4 years ago)
- Default Branch: master
- Last Pushed: 2024-10-18T02:59:07.000Z (over 1 year ago)
- Last Synced: 2024-12-07T22:12:22.327Z (over 1 year ago)
- Topics: classification, machine-learning, tensorflow, tensorflow2
- Language: Jupyter Notebook
- Homepage:
- Size: 11.9 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
## Hello World! 
🚀 Welcome to my git repo :FakeAdClicksDetection:
### Fake Ad Clicks Detection Using Tensorflow and Machine Learning
###### This project focuses on detecting fake ad clicks using advanced machine learning techniques. I applied models to classify legitimate and fraudulent clicks based on patterns and user behavior data. The system leverages data preprocessing, model training, and evaluation to optimize the detection process for real-time applications.
###### To identify the best model with the highest performance, various machine learning algorithms were evaluated. The models were fine-tuned using class weighting to account for data imbalance, resulting in improved detection accuracy for fraudulent clicks.
###### This project demonstrates how combining cutting-edge ML techniques with user-friendly interfaces can transform document interactions.
###### To run notebook - Please make sure to have the csv file in the same location or change the path.
###### To run the production code - Just run "python main.py"
###### Projects details are as follows
###### configs :- In this module I defined everything that can be configurable and can be changed in future.
###### dataloader :- I defined all dataloading, preprocessing steps in this module
###### executor :- This module is responsible for training the model.
###### evaluation :- Kept empty for now. Mainly responsible for evaluating the model.
###### model :- This module only contains the actual deep learning / machine learning code.
###### Experiments :- This directory contains the notebook which I have used for experiments. It is the same notebook attached in this email.
###### ops :- Empty as of now. Could be used in future for operational related operations
###### utils :- Utility functions that are used in more than one place.
###### saved_models :- Saved the model after training.
###### main.py :- The entry point of the project.
#### Scope for enhancement are as follows.
1- We may need to use a weighted class for training so that we can achieve a high score.
2- We could use docker containers while productionizing the model.
---
### Languages and Tools
### Languages
**Python**:
**JavaScript**:
**SQL:**
### Tools
**Software Development:**
**SDLC:**
  
**Software Engineering:**
[](https://github.com/Quananhle)
[](https://github.com/Quananhle)
**Version Control:**
---
---
Connect with me
 Â
 Â
 Â
