https://github.com/computingvictor/deeplearning_practices
Evaluation Practices done for the Deep Learning subject
https://github.com/computingvictor/deeplearning_practices
cifar10 cifar100 cunef data-science deep-learning jupyter neural-network
Last synced: 7 months ago
JSON representation
Evaluation Practices done for the Deep Learning subject
- Host: GitHub
- URL: https://github.com/computingvictor/deeplearning_practices
- Owner: ComputingVictor
- Created: 2023-02-22T13:17:22.000Z (about 3 years ago)
- Default Branch: main
- Last Pushed: 2023-02-22T14:04:37.000Z (about 3 years ago)
- Last Synced: 2025-02-16T21:41:38.347Z (about 1 year ago)
- Topics: cifar10, cifar100, cunef, data-science, deep-learning, jupyter, neural-network
- Language: Jupyter Notebook
- Homepage:
- Size: 2.22 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Deep Learning Practices
## About the project
This repository will include the evaluable practices of the subject 'Deep Learning' belonging to the Master in Data Science at CUNEF during the academic year 2022-2023.
## Built with
- Python 3.10
- Jupyter Notebook
- Visual Code
- Tensorflow
- Sklearn
- Keras
## Content of the repository
- `Practice 1- Cifar 100`:
In this practice we will perform **different tests of neural networks**. Once 8 neural networks of different characteristics have been modeled, we will keep the one that has achieved the **best result** in the **validation set**.
Subsequently, we will study how the score affects the **variation of the batch size and the number of neurons in each layer**, plotting the results.
Finally, we will make a **comparison between different Machine Learning models and the resulting scores of each one**.
- `Practice 2- Best Score`:
In this practice we will have to obtain the highest score with the CIFAR-100 dataset. For this we will have to use the different types of networks/layers and optimizers seen in class, modifying their hyperparameters, batch sizes, etc...
## Contact
Víctor Viloria Vázquez -
Project Link:
LinkedIn -