Failure Forecasting in Low Cost Sensors using Deep Time Series Models

This repository consists of the codebase for the implementation of the paper Failure Forecasting in Low Cost Sensors using Deep Time Series Models

Contents

• datasets directory consisting of the sensors data.
• models directory consisting of the saved models.
• batches.py file containing custom batch generator.
• preprocess.py module for preprocessing the data.
• train.py module to train model of choice on the data.
• test.py module to test the trained model.
• utils.py module contating the utility functions.

Requirements

• Requires anaconda

Instructions

• Create a python 3.10 environment using anaconda →

conda create env -n failurepred python=3.10

• Activate the environment →

conda activate failurepred

• Run the following command to install the dependencies →

pip install -r requirements.txt

• Use the following code to train the model →

python train.py -m <model_name> -t <test_type> -trb <train_balance_mode> -teb <test_balance_mode>

• Run the following code to get more details on the available options and default value →

python train.py --help

• Use the following code to test the model →

python test.py -m <model_name> -t <test_type> -trb <train_balance_mode> -teb <test_balance_mode>

• Run the following code to get more details on the available options and default value →

python test.py --help