Assignment 2 - IN5460

Federated Learning with RNNs and LSTMs

Felix Krause, Johannes Spies, Sina Henning

Mandatory assignment 2 for Artificial Intelligence for Energy Informatics.

Assignment | Deadline 12.11.2023 kl 23:59

Setup

Environment

Set up virtual environment

python -m venv venv

Activate custom virtual environment

source venv/bin/activate

Install required packages

pip install -r requirements.txt

Add environment to Jupyter Notebook

python -m ipykernel install --user --name=venv

Data

Place Dataset.csv: in a directory called "data"

Attention! There are two links to datasets in the docs. We are using the one which can be found on zenodo.org

Run the preprocessing notebook to create the preprocessed datasets used later.

Run via jupyter notebook

Simply execute the respective notebooks:

• task1.1_prediction_RNN.ipynb
• task1.2_prediction_LSTM.ipynb
• task2.1_2.2_classification_RNN.ipynb
• task2.3_classification_LSTM.ipynb

Run via shell

Adapt main.py to your needs and execute in project root:

python -m src

Useful tutorials

• lstm example a jupyter notebook with a time series prediction example using lstm
• rnn vs lstm vs gru code architectures of the three different models