Predicting demand for bike rentals in the next year

Data contains rental count depending on variables such as: weather conditions, day of the year, year, etc.. Data was collected over 2 years period. Project was focused on finding accurate predictive model for the task.

Dataset is available here: https://archive.ics.uci.edu/ml/datasets/bike+sharing+dataset

To run models:

1. Open Bikes_rental_analysis(1).ipynb in Jupyter/Colab.
2. Download dataset and change paths in the code.
3. Execute.

Environment requirements:

• Jupyter/Colab
• Python 3.x
• scikit-learn
• Tensorflow 2.2.0
• Libraries:
  • Matplotlib
  • NumPy
  • Pandas
  • Graphviz

Relevant paper: Fanaee-T, Hadi, and Gama, Joao, 'Event labeling combining ensemble detectors and background knowledge', Progress in Artificial Intelligence (2013): pp. 1-15, Springer Berlin Heidelberg, [Web Link].