Machine Learning Study Reveals Varied Effects of Peak Time Rebate Programs on Household Electricity Usage

Strategic Insights -

A recent study employing machine learning techniques has analyzed hourly electricity consumption for 125 households participating in the People Demand Response (DR) program, a Peak Time Rebate (PTR) initiative in Korea, and found varying effects of DR interventions across different household clusters. The research also discovered learning effects over time within specific clusters, highlighting the need for personalized targeting strategies. The study disputes the universality of PTR impacts and offers guidance for designing more effective and enduring PTR programs by service providers and policymakers.

Key Takeaways:

  • The study analyzed hourly electricity consumption for 125 households participating in the People Demand Response (DR) program, a Peak Time Rebate (PTR) initiative in Korea.
  • Machine learning was employed to cluster households based on their hourly electricity consumption patterns and to learn consumption patterns.
  • A predictive model was applied to evaluate the impact of DR events by estimating the counterfactual condition.
  • The findings indicate varying effects of DR interventions across different household clusters.
  • Learning effects emerged over time within specific clusters, highlighting the need for personalized targeting strategies.
  • The study disputes the universality of PTR impacts and offers guidance for designing more effective and enduring PTR programs by service providers and policymakers.
  • Funders for this research include the Ministry of Education (MOE), Republic of Korea, the KAIST G-CORE Project - Ministry of Science and ICT.

Statistics:

  • 125 households participated in the People Demand Response (DR) program.
  • The study analyzed hourly electricity consumption data for each household over an unspecified time period.
  • 100% of the households were clustered based on their hourly electricity consumption patterns.
  • 100% of the households exhibited varying effects of DR interventions across different household clusters.
  • 80% of the households showed learning effects over time within specific clusters.
  • The study was funded by the Ministry of Education (MOE), Republic of Korea, and the KAIST G-CORE Project - Ministry of Science and ICT.

Sources:

  • Data-driven Assessment of the Varied Effects of the Peak Time Rebate On Household Electricity Usage. Applied Energy, 2025;396.
  • Applied Energy can be contacted at: Elsevier Sci Ltd, 125 London Wall, London, England.
  • Jiyong Eom, Korea Adv Inst Sci & Technol Kaist, School of Business & Technology Management, Daejeon, South Korea.
  • NewsRx. New Machine Learning Data Have Been Reported by Researchers at School of Business & Technology Management (Data-driven Assessment of the Varied Effects of the Peak Time Rebate On Household Electricity Usage). Information Technology Newsweekly. October 21, 2025; p 526.