This project aims to predict the probability of student visa approval based on their credentials. Utilizing data collected from students' Facebook posts about their visa interview experiences, the project employs a neural network to make predictions. The model, which is an MLP (Multi-Layer Perceptron), achieves an 85% accuracy rate on the validation set.
To set up the project, follow these steps:
-
Clone the Repository
git clone https://github.com/md-aseem/visa-approval-prediction.git cd visa-approval-prediction
-
Install Dependencies
- Ensure Python 3.10 is installed.
- Install required Python packages:
pip install -r requirements.txt
- Data Collection: Scraping students’ Facebook posts using Python and structuring data with the ChatGPT API.
- Data Analysis: Analyzing the data using seaborn for insights.
- Data Modeling: Implementing and iterating over neural networks and decision trees; choosing MLP based on F-1 score.
- Model Deployment: The model is deployed using Gradio and hosted on Google Cloud Run.
The deployed model can be accessed at muhammadaseem.com/visaprobability. Over 1000 students have used this service, with feedback indicating its utility in improving visa approval chances.
Distributed under the MIT License. See LICENSE
for more information.
Muhammad Aseem - [email protected]
Project Link: https://github.com/md-aseem/visa-approval-prediction-model