{"id":27133149,"url":"https://github.com/kasraskari/cifar-10","last_synced_at":"2026-04-28T16:03:06.799Z","repository":{"id":286574158,"uuid":"961826351","full_name":"KasrAskari/CIFAR-10","owner":"KasrAskari","description":"Classifying CIFAR-10 images (CNN, KerasTuner CNN, Wide-and-Deep)","archived":false,"fork":false,"pushed_at":"2025-04-29T08:03:55.000Z","size":29,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"main","last_synced_at":"2025-06-01T00:02:37.489Z","etag":null,"topics":["cifar-10","deep-learning","jupyter-notebook","keras","keras-tensorflow","neural-network","python","tensorflow","tuning"],"latest_commit_sha":null,"homepage":"","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/KasrAskari.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","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,"zenodo":null}},"created_at":"2025-04-07T08:19:45.000Z","updated_at":"2025-05-12T11:41:57.000Z","dependencies_parsed_at":"2025-04-29T09:21:13.771Z","dependency_job_id":"5344f4fd-dae6-43ab-a786-15765fc94cff","html_url":"https://github.com/KasrAskari/CIFAR-10","commit_stats":null,"previous_names":["kasraskari/cifar-10"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/KasrAskari/CIFAR-10","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KasrAskari%2FCIFAR-10","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KasrAskari%2FCIFAR-10/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KasrAskari%2FCIFAR-10/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KasrAskari%2FCIFAR-10/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/KasrAskari","download_url":"https://codeload.github.com/KasrAskari/CIFAR-10/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/KasrAskari%2FCIFAR-10/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":32387923,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-28T14:34:11.604Z","status":"ssl_error","status_checked_at":"2026-04-28T14:32:37.009Z","response_time":56,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"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":["cifar-10","deep-learning","jupyter-notebook","keras","keras-tensorflow","neural-network","python","tensorflow","tuning"],"created_at":"2025-04-07T23:19:34.348Z","updated_at":"2026-04-28T16:03:06.763Z","avatar_url":"https://github.com/KasrAskari.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# 🧠 CIFAR-10 Deep Learning Project\n\nA deep dive into image classification using the CIFAR-10 dataset. This project implements and compares multiple deep learning architectures — including a standard CNN, a hyperparameter-tuned CNN (via KerasTuner), and a custom Wide-and-Deep model — to evaluate performance on classification tasks using **Accuracy**, **F1-score**, and **ROC-AUC**.\n\n---\n\n## 📦 Overview\n\nCIFAR-10 is a classic image classification dataset containing **50,000 32×32 color images** across **10 categories** (airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck).  \nYou can find more about the dataset [here](https://keras.io/api/datasets/cifar10/).\n\nThis project includes:\n\n- 🔹 Data preprocessing and splitting (85% train/val, 15% test)\n- 🔹 A standard deep CNN with up to 5 hidden layers\n- 🔹 Hyperparameter tuning via `KerasTuner` using `RandomSearch`\n- 🔹 A Wide-and-Deep model combining shallow and deep features\n- 🔹 Metric evaluation in tabular format (Accuracy, F1, ROC-AUC)\n\n---\n\n## 🚀 Quick Start\n\n### ✅ Prerequisites\n- Python 3.8+\n- `pip` package manager\n\n### 📥 Installation\n\n1. **Clone the repo:**\n   ```bash\n   git clone https://github.com/KasrAskari/CIFAR-10\n   cd CIFAR-10\n   ```\n\n2. **(Optional) Create a virtual environment:**\n   ```bash\n   python -m venv venv\n   source venv/bin/activate        # On Linux/Mac\n   venv\\Scripts\\activate           # On Windows\n   ```\n\n3. **Install dependencies:**\n   ```bash\n   pip install -r requirements.txt\n   ```\n\n---\n\n## 📚 Dependencies\n\nMain libraries used:\n\n- `numpy==1.26.4`\n- `tensorflow==2.16.1`\n- `scikit-learn==1.4.2`\n- `keras-tuner==1.4.7`\n- `pandas==2.2.2`\n\n(See `requirements.txt` for full list.)\n\n---\n\n## 📊 Results\n\nThe following table compares performance across models:\n\n```\nResults Table:\n   Dataset     Deep Accuracy  Deep F1  Deep ROC-AUC  Tuned Accuracy  Tuned F1  Tuned ROC-AUC  WideDeep Accuracy  WideDeep F1  WideDeep ROC-AUC\n0  Train       0.85           0.84     0.98          0.87            0.86      0.99           0.86               0.85         0.98\n1  Validation  0.75           0.74     0.92          0.78            0.77      0.94           0.76               0.75         0.93\n2  Test        0.73           0.72     0.91          0.76            0.75      0.93           0.74               0.73         0.92\n```\n\n📌 *Values are sample outputs. Final results depend on training runs.*\n\n---\n\n## 📝 Notes\n\n- Images are normalized to `[0, 1]`.\n- Labels are one-hot encoded for use with `CategoricalCrossentropy`.\n- Model training time may vary depending on hardware (CPU vs. GPU).\n- For best performance, ensure TensorFlow is CUDA-enabled if using a GPU.\n\n---\n\n## 🤝 Contributing\n\nPull requests are welcome! Feel free to fork the project, open issues, or suggest improvements.\n\n---\n\n## 📄 License\n\nThis project is licensed under the [MIT License](LICENSE).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkasraskari%2Fcifar-10","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fkasraskari%2Fcifar-10","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fkasraskari%2Fcifar-10/lists"}