cascade-multivariate-paper

How to run

Install dependencies

First, you need to install the dependencies:

python -m pip install -r requirements.txt

Data preprocessing

Then, you can run the preprocessing of the data script:

python gen_data.py --data=<dataset> --hours=1 --p_steps=0

• --data is the dataset to use. The options are air_quality, traffic, energy, power, parking, room, solar, kolkata, turbine, joho, electricity, iot and wind.
• --p_steps is the number of previous time steps to add as features to the final dataset.

You can also run the preprocessing script for all the datasets, with the paper's parameters with the following bash script:

/bin/bash gen_data.sh

Training

For training the model, you can run:

python training_pipeline.py  --data=<dataset> --logging=info --ii=<int>

• --data is the dataset to use. The options are air_quality, traffic, energy, power, parking, room, solar, kolkata, turbine, joho, electricity, iot and wind.
• --logging is the logging level. The options are debug, info, warning, error and critical. The default is info.
• --ii is the number of the importance iterations of the experiment. If it's 0 then feature importance is not calculated. The default is 0.

Dataset descriptions

Traffic Dataset

Download & information about the dataset

Power Consumption of Turkey (2016-2020)

Download & information about the dataset

Energy consumption Dataset

Download & information about the dataset

Parking Occupancy Dataset

Download & information about the dataset

Temperature and Humidity of Kolkata from 2015-2020

Download & information about the dataset

Solar Generation in Italy Dataset

Download & information about the dataset

Wind Turbine Texas Dataset

Download & information about the dataset

Hourly load data of the power supply company of the city of Joho

Download & information about the dataset

Temperature Readings : IOT Devices

Download & information about the dataset

NYC Taxi Traffic

Download & information about the dataset