Anomaly Detection in BACnet systems
Meeting 21
Title: Meeting to decide on preprocessed data
Date: 03/23/2020
Time: 11:30 am
Location: Zoom Conference Call
Attendees: Sofian Ghazali, Rahul Balamurugan, Zahid Kamil
Absences: None
Action Items: Go over the progress been done
Meeting Agenda:
-
Identify what needs to be done in Machine Learning
-
Identify Machine Learning output and decipher patterns in BACnet data
Discussion and Notes:
The meeting focused on understanding how to interpret the output from the different Machine Learning models we considered from Recurrent Neural Networks to Random Forest to clustering analysis. Therefore on our categorical data and normal data we have to continue to try different forms of preprocessing data in order for the anomaly detection to take place
Next meeting action items:​
-
Decide on how to better preprocess the data so that the Machine Learning model can understand
Meeting Adjournment: 12:00 pm
Meeting Minutes prepared by: Zahid Kamil