The goal of this project is to build an educational website that allows users to:
- Explore different machine learning algorithms.
- Gain beginner-friendly insights without learning programming.
- Learn machine learning concepts in a fun and engaging way.
Many beginners struggle with machine learning due to coding requirements and complex concepts. This platform allows users to experiment and explore ML algorithms interactively, without needing prior programming knowledge.
- Explore existing machine learning algorithms.
- Add new algorithms with minimal setup.
- Upload Pickle files for models.
- Upload CSV files for datasets.
- Train and test models.
- Receive notifications for updates.
- Subscribe/unsubscribe for algorithm updates.
.jpg)
.jpg)
- Follows MVC (Model-View-Controller) architecture.
- Uses Simple Factory Pattern for adding new algorithms.
- Uses Observer Pattern for notifications to learners.
- PyCharm: Development IDE for easy project management.
- Django: Backend framework implementing MVC architecture.
- Scikit-learn: For creating, training, and testing ML algorithms.
- MongoDB Atlas: Stores unstructured data as JSON objects.
- Docker: Containerized environment for easy setup and consistent development.
- The original project relied on a MongoDB M0 cluster with a snapshot.
- Recovery steps were attempted: restoring to an M10 cluster, exporting via
mongodump/mongorestore. - Issue: The snapshot used an old MongoDB version with
.wtfiles.- Modern MongoDB clusters no longer support this version.
- Archived versions for download are not available.
.wtfiles cannot be directly converted to JSON.
- Result: Original dataset is lost. The project cannot be fully functional as originally intended.
- The application structure, code, and functionality are preserved and can be adapted to new datasets or databases.
- TODO: Reach out to MongoDB priority team, and seek help. If not possible, manually recreate the whole dataset.