"Smart Home Energy Consumption Dataset for AI-Driven Residential Load Forecasting and Demand Response"

"This Descriptor article presents the Appliance-Level Household Energy Consumption Dataset, a high-quality, curated resource designed to accelerate AI research in power and energy systems. The dataset comprises 100,000 records from 500 anonymized smart homes, capturing detailed energy consumption for nine distinct appliance types. Each record is enriched with critical contextual features including precise timestamps, outdoor temperature, household size, and seasonal information. The data undergoes rigorous quality assurance with zero missing values and zero duplicate records. Comprehensive machine learning benchmarks\u2014from linear regression to gradient boosting and neural networks\u2014demonstrate strong predictive performance (R\u00b2 > 0.75) and validate the dataset's utility for AI model training. The dataset is taken from Kaggle and released under Creative Commons Public Domain CC0 by 1.0 Universal licence to promote open science and reproducibility and take care about data privacy and ethics. Primary use cases include appliance-level load forecasting, occupancy analysis, demand response modeling, and AI algorithm benchmarking for smart grid applications. "