Predictive Process Monitoring Dashboard for Emulating LSTM Models over Business Processes and its Qualitative Evaluation

The project is developed as the part of Master Thesis Project at Eindhoven University of Technology, under the guidance of Dr. Dirk Fahland.

The thesis report can be found here.

The base code of this project is taken from GenerativeLSTM v1.1.0 which is based on Camargo et al. paper Learning Accurate LSTM Models of Business Processes.

Streamlit has been used to create the dashboard, and a trimmed down version of this project is kept in Streamlit based Predictive Process Monitoring powered by LSTM for hosting it on freeware servers.

Getting Started

These instructions will help you set up a development and testing copy of the project on your local machine.

Prerequisites

• To run this code, first install Anaconda on your system, don't foget to check the PATH while installation.
• Change the directory to desired location where you would like to clone the repository, and then clone it.
• Create Conda virtual Environment using conda create -n <env name>
• Activate the Virtual Env : conda activate <env name>
• Install Python 3.7 using conda install python=3.7
• Then install the packages required from the requirement.txt file using pip install -r requirements.txt

Running the script

Once you've established the environment, you may run the dashboard using the streamlit run dashboard.py from the root directory.