| Autor: |
Zhang, Hai-Tao, Wang, Zhiyue, Liu, Xingjian, Zhou, Wei, Ding, Yizhou, Li, Yuanzheng, Hu, Jiabing |
| Zdroj: |
IEEE Journal of Emerging and Selected Topics in Industrial Electronics; 2024, Vol. 5 Issue: 2 p652-661, 10p |
| Abstrakt: |
To extract the household attribute information from the large volume of smart meter data, this study proposes a resident characteristics estimator. Such an estimator enables energy suppliers to provide personalized services whereas to assist customers to reduce energy consumption. By leveraging the potential connections among different characteristics, a deep convolutional neural network-based cross-task transfer learning scheme is designed, which makes full use of the knowledge learned from one characteristic (such as retirement status)-based classification to estimate another relevant characteristics (such as age). Extensive experiments are conducted on the Irish dataset with 4232 households to substantiate the superiority of the proposed scheme compared with conventional deep convolutional neural networks-based learning methods. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
