📌 Description This repository contains Jupyter Notebooks developed as supplementary material for the article "Introduction to Neural Networks for Physicists", published in the Arxiv. Created by Gustavo Café and Gubio Gomes, these notebooks aim to enhance the understanding of the concepts discussed in the article by providing practical and interactive examples. Specifically, they explore different neural network approaches applied to the dynamics of a simple pendulum, illustrating their potential in computational physics.
📁 Notebooks
| 📌 File | 📖 Description |
|---|---|
🟢 01-Perceptron-iris.ipynb |
Introduction to the Perceptron algorithm applied to the Iris dataset. |
🟠 02-Perceptron-Regressao.ipynb |
Using the Perceptron for regression problems. |
🔵 03-Perceptron-iris.ipynb |
Another implementation of the Perceptron on the Iris dataset with specific adjustments. |
⚙️ 04-Exemplo 1 Pendulo.ipynb |
Determine the constant for OHS (Simple Harmonic Oscillation) when modeling a pendulum using neural networks. |
🔥 05-Exemplo 2 OHS.ipynb |
Solve diferential equation of OHS with method call Physics informed neural network. |
🏗️ ./Exemplo 3-Autoencoder |
Implementation of an Autoencoder for compression and representation learning. |
🤖 ./SINDyAutoencoder |
Use of Sparse Identification of Nonlinear Dynamics (SINDy) with Autoencoders. |
📊 Other Files
- 🖼️
SINDyAutoencoder_ValidationData_PT.png- Validation figure for the SINDyAutoencoder model.
1️⃣ Clone the repository
git clone https://github.com/Coffee4MePlz/Notebooks_NN_Physics.git2️⃣ Install dependencies (if needed)
pip install -r requirements.txt3️⃣ Run the notebooks
- Open Jupyter Notebook:
- Navigate to the desired notebook and execute the cells.
- Python 3.x
- Jupyter Notebook
- Libraries: NumPy, TensorFlow, SciPy, Matplotlib (details in
requirements.txt)
If you have any questions or suggestions, feel free to reach out:
📧 Email: gcaf0125@uni.sydney.edu.au
🔗 LinkedIn: My Profile
⭐ If you find this useful, don't forget to star the repository! ⭐