Scholarlink

The motivation for employing time series models stems from their superior performance compared to analysts’ long-term wage forecasts, especially for small firms. This study underscores the challenges financial commentators face in making accurate predictions. The primary objective is to evaluate the effectiveness of time series models in forecasting household electricity consumption. This paper presents an advanced data analytics approach using a Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) hybrid network. Data were collected at a one-minute sampling rate over nearly four years, focusing on the Global Active Power variable. The CNN-LSTM hybrid network leverages the strengths of both CNNs and LSTMs to enhance prediction accuracy. The results demonstrate the network's potential in providing reliable forecasts of household energy consumption, offering valuable insights for efficient energy management and planning. Exploratory analysis using line plots and histograms is conducted to understand data distributions and usage patterns over the four-year period.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Advanced-Data Analytics for Household Power Consumption Forecasting Using CNN-LSTM Hybrid Network

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Chandani Sharma,
  • Smriti Sachan,
  • Sohini Chowdhury,
  • Rini Saxena,
  • Archana Saxena,
  • Laith H. Jasim

摘要

The motivation for employing time series models stems from their superior performance compared to analysts’ long-term wage forecasts, especially for small firms. This study underscores the challenges financial commentators face in making accurate predictions. The primary objective is to evaluate the effectiveness of time series models in forecasting household electricity consumption. This paper presents an advanced data analytics approach using a Convolutional Neural Network-Long Short-Term Memory (CNN-LSTM) hybrid network. Data were collected at a one-minute sampling rate over nearly four years, focusing on the Global Active Power variable. The CNN-LSTM hybrid network leverages the strengths of both CNNs and LSTMs to enhance prediction accuracy. The results demonstrate the network's potential in providing reliable forecasts of household energy consumption, offering valuable insights for efficient energy management and planning. Exploratory analysis using line plots and histograms is conducted to understand data distributions and usage patterns over the four-year period.