TITLE:
Machine Learning: An Overview (Part Two)
AUTHORS:
Mohd Izhan Mohd Yusoff
KEYWORDS:
Nonintrusive Load Monitoring (NILM), Combinatorics, Observation Matrix, Pentagon
JOURNAL NAME:
Open Journal of Modelling and Simulation,
Vol.14 No.2,
April
24,
2026
ABSTRACT: NILM (Nonintrusive Load Monitoring) or appliance recognition software that uses algorithms, namely Hidden Markov Model (HMM) and Factorial HMM, to detect changes in electricity values entering a building, thus inferring which appliances are used and their individual energy consumption. In this paper, a “simple” technique is introduced that uses combinatorics on a specially designed observation matrix to detect hidden appliances and their electricity usage by dissecting the building’s total electricity usage.