Daily Arxiv

This is a page that curates AI-related papers published worldwide.
All content here is summarized using Google Gemini and operated on a non-profit basis.
Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.

Hybrid Approach for Electricity Price Forecasting using AlexNet and LSTM

Created by
  • Haebom

Author

Bosubabu Sambana, Kotamsetty Geethika Devi, Bandi Rajeswara Reddy, Galeti Mohammad Hussain, Gownivalla Siddartha

Outline

This paper proposes a novel hybrid model that combines AlexNet and LSTM algorithms to improve the accuracy of electricity price prediction. Existing RNN and ANN models have difficulty in processing foreign exchange time series data and often fail to accurately predict prices, but the hybrid model proposed in this paper significantly improves the prediction accuracy by considering external variables such as demand, temperature, amount of sunshine, and rainfall. By utilizing the excellent feature extraction ability of AlexNet and the time series pattern learning ability of LSTM, methods such as min-max scaling and time window are applied to predict future electricity prices, and the accuracy is 97.08%, which is higher than the existing RNN (96.64%) and ANN (96.63%) models.

Takeaways, Limitations

•
Takeaways:
◦
We show that a hybrid model combining AlexNet and LSTM is effective in improving electricity price prediction accuracy.
◦
It is suggested that prediction accuracy can be improved by considering external variables (demand, temperature, amount of sunlight, rainfall, etc.).
◦
Achieved higher prediction accuracy (97.08%) than existing RNN and ANN models.
•
Limitations:
◦
Further validation of the generalization performance of the model presented in the paper is needed.
◦
Experimental results on various datasets are limited.
◦
Lack of analysis of the model's computational complexity and running time.
◦
There is a lack of specific evidence and detailed explanation for the high accuracy of 97.08%.
👍
Made with Slashpage