Big Data Analytics Extension Course

• Theme: Neural Networks for Reliability

• Professor: Enrique Andres López Droguett (UCLA)

• Workload: 36 h (Jan. and Feb.2023)

• Folder with datasets: datasets. Download the file Reboiler_Raw Data_With Timestamps.csv and move it to the downloaded/cloned folder (big_data_analytics_extension_course)

Instructions (for Mac and Linux users with Docker installed and available GPU)

Move to the docker folder

cd docker

Set execution permission for all users

chmod a+x run_image.bash

Use the generated kernel link to access the jupyter notebook

Run and study the jupyter notebook reboiler_binary_classifier.ipynb.

Instructions (for Mac and Linux users without Docker installed and/or without GPU, or if you have all necessary configuration to run tensorflow with GPU (NVIDIA Toolkit, cuDNN, TensorRT...) and don't need Docker):

First, install the requirements (preferably in a virtual environment):

python3 -m venv virtual-env

Activate the environment:

source virtual-env/bin/activate

Install the requirements:

pip install -r requirements.txt

Run and study the jupyter notebook reboiler_binary_classifier.ipynb.

Instructions (for Windows users):

Repeat the procedure above using conda instead of pip. It's easy to find instructions on how to use conda on the internet.

Paper

A paper registering the methodology and results is available in the paper folder. It is necessary to have latex installed to compile the file main.tex.