An Improved Gradient Enhancement Regression Tree Method to Evaluate the Innovation Capability of Energy Companies

Abstract

In recent years, with the introduction of the innovation-driven development strategy, audit departments across the country have actively responded by arranging audit experts to assess the innovation capabilities of enterprises based on their information. They have also studied the impact of implementation on innovation indicators in energy enterprises, aiming to accurately implements to support enterprises and drive regional development. Traditional manual evaluation methods are inefficient and prone to human interference. By using a gradient boosting regression tree model to construct a scoring prediction model, instead of manual evaluation methods, both accuracy and efficiency can be ensured. Experimental results show that this prediction model outperforms other models such as random forest regression and can guarantee prediction accuracy.