Session Category : Advanced Analytics
Improving customer satisfaction for distribution system reliability is possible through Machine Learning and Data Analytics. Predicting the likelihood for an individual customer to complain, is achievable, by applying data science to the abundant of information available through AMI and historically collected data. This session will explore DTE Energy’s machine learning lifecycle with forecasting frequent outage complaints, and would be applicable to additional use cases like customer outage experience, capital investment, work scheduling, emergent work & resources, etc.
This presentation will assist with:
- Leveraging AMI with outage management system data to predict individual likelihood to complain
- Hyper Parameter Tuning to utilize best in class models
- Predicting Complaint Likelihood
- Utilizing visualization capabilities to allow for a proactive approach
- Integrating Preventative Solutions
- Model optimization and continuous improvement
By the end of this presentation, our goal is to empower you to drive change that will better your organization with utilizing machine learning.