- genre: lightning talk
- title: Data-Driven Demand Estimation in Electrification: A Machine-Learning Approach to Enhance Rural Energy Systems
- presenter: @Ale_onori
- description: This lightning talk addresses the critical issue of achieving universal energy access through a dual approach involving comprehensive guidelines and a data-driven machine-learning framework. Initially, it outlines a detailed methodology based on extensive literature review, providing guidelines for developing data-sharing platforms in energy access. This includes a proposed architecture to support the collection of conjoint socio-economic and time-series data. Building upon this foundation, the talk then delves into the development of a machine-learning framework for estimating appliance adoption patterns. This framework leverages a novel open-access database, populated with socio-economic information and appliance data collected from various public and private sources. A structured logistic regression analysis is employed to discern the nexus between socio-economic factors and appliance adoption, aiming to identify the most relevant indicators for demand estimation.