This repository contains Python code for predicting the temperature 1 hour into the future using two different Recurrent Neural Network (RNN) models: one with simple vanilla RNN cells and the other with Gated Recurrent Unit (GRU) cells. The models are trained and evaluated using the "jena_climate_2009_2016" dataset.
The temperature prediction models are trained and evaluated using the "jena_climate_2009_2016" dataset. This dataset contains weather measurements recorded at the Weather Station in Jena, Germany, from 2009 to 2016.
To use this dataset, follow these steps:
-
Download the dataset from the following source link: jena_climate_2009_2016.csv.zip.
-
Extract the contents of the zip file.
-
The dataset file should be named "jena_climate_2009_2016.csv".
-
Place the "jena_climate_2009_2016.csv" file in the project's root directory.
Make sure you have the following dependencies installed:
- NumPy
- Matplotlib
- scikit-learn
- TensorFlow
- Pandas
- Seaborn
- pandas-profiling
You can install the dependencies using the following command:
pip install numpy matplotlib scikit-learn tensorflow pandas seaborn pandas-profiling-
Clone the repository to your local machine:
git clone https://github.com/Roon311/RNN-Models.git
-
Navigate to the project directory:
cd RNN-Models -
Run the Jupyter Notebook
Temperature_Prediction.ipynbto train and evaluate the RNN models for temperature prediction. -
Follow the instructions and code provided in the notebook to understand and execute the temperature prediction process using both vanilla RNN and GRU models.
-
Experiment with different hyperparameters, architectures, or preprocessing techniques to improve the model's performance.
For any questions or inquiries, please contact [s-noureldin.hamed@zewailcity.edu.eg].
Good luck with your temperature prediction models!