---
title: "Improving Wind and Solar Power Prediction with Efficient Wrapper-based Feature Selection: An Empirical Study"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.14024v1"
permalink: "https://feed.astralor.com/items/itm_41ead1c26296304dc9f7007c"
published: "2026-07-15T16:55:42.000Z"
score: 71
channel: "tech"
tags: ["CSFS","SFS","Wrapper Method","Renewable Energy Prediction","Feature Selection"]
---

# Improving Wind and Solar Power Prediction with Efficient Wrapper-based Feature Selection: An Empirical Study

## AI commentary

> This paper is valuable for researchers and engineers working on renewable energy forecasting or automated feature selection. The clustering-based acceleration idea of CSFS is simple yet effective, and the open-source code makes it easy to replicate. If you're struggling with feature selection for wind or solar prediction models, the comparative experiments in this paper can directly guide your choice of strategy. What I find most interesting is that the authors didn't chase the latest complex methods but instead solved computational efficiency with a practical engineering approach.

## At a glance

This paper addresses the lack of systematic feature selection methods in renewable energy prediction by proposing Cluster-based Sequential Feature Selection (CSFS), a novel model-agnostic wrapper method that uses clustering for efficient and reliable feature selection. Evaluated on wind turbine power curve modeling and photovoltaic power prediction, CSFS achieves predictive performance comparable to standard sequential feature selection (SFS) while reducing computational cost by an average of 21%. The paper provides open-source implementation and compares CSFS with filter-based, wrapper-based, and embedded methods, showing wrapper methods generally outperform others.

## Key points

- Proposes CSFS, a clustering-based wrapper feature selection method for renewable energy prediction.
- CSFS reduces computational cost by 21% on average while maintaining predictive performance.
- Reveals gaps in existing feature selection methods through two literature reviews and empirical comparisons.

## Full text

With rising global energy demand and growing awareness of climate change and its impacts, the share of renewable energies in the global energy mix continues to grow. Unlike conventional power generation, the output of renewable energy sources cannot be controlled as consistently due to their dependence on environmental conditions. Therefore, reliable prediction of current and future energy production is essential. In this paper, we report findings from two structured literature reviews on real-world renewable energy prediction tasks: wind turbine power curve modeling and photovoltaic power prediction. For the former, we conducted a comprehensive literature review ourselves, while for the latter, we synthesize the key findings regarding frequently selected input features based on an existing survey. Across both domains, our analysis reveals that despite the large number of available monitoring and environmental variables, only limited or unsystematic methods for feature selection exist. To address this gap, we propose Cluster-based Sequential Feature Selection (CSFS), a novel, model-agnostic, clustering-based wrapper method for automatic, efficient, and reliable feature selection in renewable energy prediction pipelines. To support reproducibility and reuse, we provide an open-source implementation of CSFS on GitHub. We empirically evaluate the proposed approach on both use cases and compare it with established feature selection techniques such as wrapper-based sequential feature selection (SFS), filter-based methods, and Random Forest's embedded feature importance. The results show that the wrapper-based methods overall provide better-performing selections of features. CSFS achieves a predictive performance comparable to SFS while reducing computational cost by an average of 21%.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_41ead1c26296304dc9f7007c