A Few-Shot LLM Framework for Extreme Day Classification in Electricity Markets
arXiv:2602.16735v1 Announce Type: new Abstract: This paper proposes a few-shot classification framework based on Large Language Models (LLMs) to predict whether the next day will have spikes in real-time electricity prices. The approach aggregates system state information, including electricity demand, renewable generation, weather forecasts, and recent electricity prices, into a set of statistical features that are formatted as natural-language prompts and fed to an LLM along with general instructions. The model then determines the likelihood that the next day would be a spike day and reports a confidence score. Using historical data from the Texas electricity market, we demonstrate that this few-shot approach achieves performance comparable to supervised machine learning models, such as Support Vector Machines and XGBoost, and outperforms the latter two when limited historical data are available. These findings highlight the potential of LLMs as a data-efficient tool for classifying
arXiv:2602.16735v1 Announce Type: new Abstract: This paper proposes a few-shot classification framework based on Large Language Models (LLMs) to predict whether the next day will have spikes in real-time electricity prices. The approach aggregates system state information, including electricity demand, renewable generation, weather forecasts, and recent electricity prices, into a set of statistical features that are formatted as natural-language prompts and fed to an LLM along with general instructions. The model then determines the likelihood that the next day would be a spike day and reports a confidence score. Using historical data from the Texas electricity market, we demonstrate that this few-shot approach achieves performance comparable to supervised machine learning models, such as Support Vector Machines and XGBoost, and outperforms the latter two when limited historical data are available. These findings highlight the potential of LLMs as a data-efficient tool for classifying electricity price spikes in settings with scarce data.
Executive Summary
This study proposes a novel few-shot classification framework leveraging Large Language Models (LLMs) to predict spikes in real-time electricity prices. By aggregating system state information into statistical features, the model achieves performance comparable to traditional supervised machine learning models, with a notable advantage when faced with limited historical data. The findings demonstrate the potential of LLMs as a data-efficient tool in electricity market analysis, where data scarcity is common. The approach presents a promising solution for real-time forecasting and decision-making in electricity markets.
Key Points
- ▸ The proposed few-shot LLM framework aggregates system state information into statistical features for electricity price spike prediction.
- ▸ The model achieves comparable performance to traditional supervised machine learning models, such as Support Vector Machines and XGBoost.
- ▸ The LLM framework outperforms traditional models when limited historical data are available, highlighting its data-efficient capabilities.
Merits
Strength in Data Efficiency
The few-shot LLM framework demonstrates the potential to effectively address data scarcity in electricity market analysis, a common challenge in this domain.
Adaptability to Real-Time Data
The framework's ability to incorporate real-time system state information enables timely decision-making in electricity markets.
Demerits
Model Dependence on Training Data
The performance of the LLM framework is contingent upon the quality and quantity of the training data, which may not always be readily available.
Limited Generalizability
The study's focus on the Texas electricity market may limit the framework's applicability to other markets with different characteristics.
Expert Commentary
The proposed few-shot LLM framework presents a promising solution for electricity price spike prediction, addressing a critical challenge in electricity market analysis. The study's findings demonstrate the potential of LLMs in data-efficient applications, where adaptability and real-time decision-making are essential. However, the framework's model dependence on training data and limited generalizability to other markets should be carefully considered. Further research is needed to explore the framework's applicability to different market settings and to develop more robust and transferable LLM models for electricity market analysis.
Recommendations
- ✓ Future studies should investigate the framework's performance in other electricity markets with different characteristics, to assess its generalizability and adaptability.
- ✓ Researchers should explore the use of alternative LLM architectures and training methodologies to improve the framework's robustness and transferability.
