{"id":19426406,"url":"https://github.com/Nishat-Ahmad/Youtube-Projects","last_synced_at":"2025-04-24T17:31:05.469Z","repository":{"id":260925120,"uuid":"882726336","full_name":"Nishat-Ahmad/Youtube-Projects","owner":"Nishat-Ahmad","description":"All of the projects that I have went through not made by myself.","archived":false,"fork":false,"pushed_at":"2025-01-07T11:21:47.000Z","size":19864,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-02-25T05:16:04.299Z","etag":null,"topics":["classification","digit","mnsit-database","neural-network"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/Nishat-Ahmad.png","metadata":{"files":{"readme":null,"changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2024-11-03T15:48:53.000Z","updated_at":"2025-01-07T11:21:50.000Z","dependencies_parsed_at":"2025-01-07T12:38:08.656Z","dependency_job_id":null,"html_url":"https://github.com/Nishat-Ahmad/Youtube-Projects","commit_stats":null,"previous_names":["nishat-ahmad/digit-classification","nishat-ahmad/youtube-projects"],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Nishat-Ahmad%2FYoutube-Projects","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Nishat-Ahmad%2FYoutube-Projects/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Nishat-Ahmad%2FYoutube-Projects/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/Nishat-Ahmad%2FYoutube-Projects/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/Nishat-Ahmad","download_url":"https://codeload.github.com/Nishat-Ahmad/Youtube-Projects/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":250674259,"owners_count":21469186,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["classification","digit","mnsit-database","neural-network"],"created_at":"2024-11-10T14:07:32.099Z","updated_at":"2025-04-24T17:31:05.452Z","avatar_url":"https://github.com/Nishat-Ahmad.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Neural Network from Scratch in Python\r\n\r\nThis project is an implementation of a neural network from scratch using only `numpy`. \r\nThe goal is to classify handwritten digits from the MNIST dataset, an introductory machine \r\nlearning problem that involves recognizing digits from 28x28 pixel grayscale images.\r\n\r\nThis implementation does not use high-level machine learning libraries like TensorFlow or Keras, \r\nfocusing instead on understanding the foundations of neural networks by implementing the necessary \r\nmath and processes from the ground up.\r\n\r\n## Overview\r\n\r\nThe neural network has:\r\n- **Three Layers**:\r\n  - Input layer with 784 nodes (one for each pixel in the 28x28 image)\r\n  - Hidden layer with 10 nodes\r\n  - Output layer with 10 nodes, each representing a digit from 0 to 9\r\n- **Activation Functions**:\r\n  - ReLU (Rectified Linear Unit) for the hidden layer\r\n  - Softmax for the output layer to interpret predictions as probabilities\r\n\r\nThis neural network implementation includes:\r\n1. **Forward Propagation**: To compute predictions\r\n2. **Backpropagation**: To calculate gradients and optimize weights and biases\r\n3. **Gradient Descent**: To iteratively minimize the loss function by adjusting weights and biases\r\n\r\n## Key Concepts\r\n\r\n- **Activation Functions**: ReLU and Softmax help the network learn complex patterns beyond simple linear transformations.\r\n- **Cost Function**: Measures the difference between the network’s prediction and the actual label.\r\n- **Learning Rate**: Controls the size of updates to the weights and biases during gradient descent.\r\n\r\n## Dataset\r\n\r\nThe MNIST dataset provides tens of thousands of labeled examples of handwritten digits in low-resolution grayscale format. \r\nEach image is represented as a 784-dimensional vector (28x28 pixels), with pixel values between 0 (black) and 255 (white).\r\n\r\n## Code Structure\r\n\r\n- `initialize_parameters`: Initializes weights and biases for the layers.\r\n- `forward_propagation`: Computes the output predictions based on the current weights and biases.\r\n- `back_propagation`: Calculates gradients for updating weights and biases to reduce error.\r\n- `gradient_descent`: Orchestrates the learning process by iterating through forward and backward passes, updating parameters on each iteration.\r\n\r\n## Installation\r\n\r\n1. Clone the repository:\r\n   ```bash\r\n   git clone https://github.com/Nishat-Ahmad/Digit-Classification.git\r\n   cd Digit-Classification\r\n\r\n2. Install dependencies:\r\n    ```bash \r\n    pip install numpy pandas matplotlib\r\n\r\n3. Obtain the MNIST dataset and place it in the project directory.\r\n   \r\nRunning the Project\r\nTo train and test the neural network:\r\nPreprocess the dataset (e.g., shuffle and split into training and validation sets).\r\nRun the script:\r\n  ```bash\r\n  python main.py\r\n  ```\r\nCredits\r\nThis project was inspired by Samson Zhang's tutorial on YouTube (https://www.youtube.com/watch?v=w8yWXqWQYmU\u0026t=541s),\r\nwhich provided the framework for implementing a neural network without machine learning libraries.\r\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FNishat-Ahmad%2FYoutube-Projects","html_url":"https://awesome.ecosyste.ms/projects/github.com%2FNishat-Ahmad%2FYoutube-Projects","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2FNishat-Ahmad%2FYoutube-Projects/lists"}