A open-source repository hosting some fun projects that you can work on to improve your ML/DL skills. We will continue to add more projects in the future for all skill levels, ranging from beginner to advanced as our initiative of promoting learning and open source contribution among all.
Here's a list of the showcased projects in this repo:
This project deals with the age old problem in optical character recognition— Dealing w/ noise in the image data that can lead up to misrepresentation and inaccuracies during inference. Here, we implement an AutoEncoder model using TensorFlow and Keras that eliminates noise/distortions within the image data for better OCR operation.
A Machine Learning project that makes the use of the KNN algorithm to recommend books to the users based off of the average ratings of the books in the data, the language they are written in, and the rating for that book.
The Deep Space comprises of innumerable celetial bodies— planets, stars, galaxies, asteroids, etc. As a result, it is not possible to label each of these celetial bodies via a more traditional manual method. This is where machine learning shines, which allows Scientists to label a celestial body based on a variety of features like its gradient and standard deviation in a 2 dimensional space, etc. In this project, some models have been implemented, based on the same principles for classification of celestial bodies based on their features.
Each year, hundreds of thousands of women lose their lives to cervical cancer, especially in developing countries where people often neglect/can't afford regular checkups and pap tests. This beginner-level project aims at an early detection of the risk of cervical cancer using different machine algorithms.
The objective of this project was to predict the amount of followers gained by a streamer on Twitch based on the streaming data. Different visualization and data analysis techniques were used for understanding the data as well as deriving various insights from it.
A beginner-level project that uses different machine algorithms to predict whether a student will get placed into a job via campus recruitment or not.
In order to contribute to an existing project, the following are the guidelines you must follow:
- Modeling and Optimization: For the given problem statement/dataset in a project, if your proposed solution (can be a different model architecture, or optimized set of hyperparameters for the existing model) performs better than the existing one, then your submission will be accepted.
In this case, kindly submit a separate jupyter notebook, and while raising a PR, include the screenshot of your results. Naming scheme for new jupyter notebook: <architecture_name>.ipynb
- Code Optimization: You can improve the space-time complexity of the existing code in a given project. For example- Using vectorization and broadcasting methods to eliminate loops, etc.
For this, modify the code in the existing jupyter notebook.
- Resolving Existing Issues: You can address the existing issues within the repository.
- Contributing your Personal Projects: If you have a fun personal project that you would love to showcase here, raise an issue regarding the same and we will get back to you soon.
 |
|---|
| Aman Sharma |
 |
Need help? Feel free to contact me @ amansharma2910@gmail.com
Aman Sharma, ML Lead @ DSC VIT © 2020
Contributions by amansharma2910: Noise Removal and OCR Using CNNs and Autoencoders || Cervical Cancer Risk Prediction
Contributions by AM1CODES: Book Recommender System || Campus Recruitment Analysis
Contributions by kritikashah20: Celestial Bodies' Classification
Contributions by Ani0202: Added KNN, Logistic Regression and SVM Classifiers; improved Decision Tree Classifier in Cervical Cancer Detection project