-
Notifications
You must be signed in to change notification settings - Fork 76
Open
Labels
Points: 30open-for-allAny one can work on it without getting assigned, every PR can mergeAny one can work on it without getting assigned, every PR can merge
Description
Issue: Understand the Dataset & Perform Initial Exploration
Objective
Before building any model, it is crucial to understand the dataset thoroughly.
This issue focuses on exploring the dataset structure, classes, distributions, and visual patterns in the images.
Dataset
- Name: SpaceNet – Astronomy Image Dataset
- Link: https://www.kaggle.com/datasets/razaimam45/spacenet-an-optimally-distributed-astronomy-data
Participants can download the dataset and work locally, or work on Kaggle (Recommended)/colab directly and upload the notebook.
Tasks to Perform
You must include (but are not limited to) the following:
- Identify all classes and number of samples per class and Check class balance / imbalance [Plot the required plot]
- Visualize random samples from each class
- Analyze image resolutions, formats
Submission Guidelines
- Fork the repository
- Create a folder inside
participants/named exactly as your enrollment number - Add your exploratory notebook(s) inside that folder
- Commit and push your changes
- Open a Pull Request referencing this issue
- Follow the template as in previous issues [ Issue: #<issue_number> ]
participants/
└── <your_enrollment_number>/
└── data_exploration.ipynb
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
Points: 30open-for-allAny one can work on it without getting assigned, every PR can mergeAny one can work on it without getting assigned, every PR can merge