Identifying Overlapping Successive Events Using a Shallow Convolutional Neural Network
This is the technical road map: 
Research
Cognitive Web Application based on IBM Watson API
Embed our identication algorithms to the power system control system to improve human-computer interactive interface on cloud. 
Application of Factorial Hidden Markov Model (FHMM) to Non-intrusive Load Monitoring
Course Project, Rensselaer Polytechnics Institute, Troy NY, USA,
Implement the factorial hidden markov chain model (FHMM) to disaggregate loads 
Real-time Faulted Line Localization and PMU Placement in Power Systems through Convolutional Neural Networks
The proposed method is able to locate the faulted line in the wide-area network with low obervability. The critical point is to extract features based on the physical characteristics. 
Fast Event Identification through Subspace Characterization of PMU Data in Power Systems
This research learn a dictinary of subspace to identify events efficiently 