Scholarlink

As demand for digital technologies grows, so does the impact of greenhouse gas (GHG) emissions. This paper explores innovative strategies to reduce carbon dioxide (CO2e) emissions in the product and technology industry, focusing on sustainable solutions to minimize environmental impact during product processes and transformations. The integration of artificial intelligence and Key Performance Indicators (KPIs) is significant within industrial sectors. This study proposes a machine learning-based approach to predict KPIs aimed at minimizing GHG emissions across product processes and transformations. A dataset obtained from the Environmental Protection Agency (EPA) portal was used; feature selection is performed using the Recursive Feature Elimination (RFE) method. We extract relevant KPIs from the dataset and mathematically represent these KPIs. We train six machine learning (ML) classifiers: Random Forest (RF), Decision Tree (DT), K-nearest neighbors (K-NN), Gradient Boosting (GB), Adaboost, and Light Gradient Boosting Machine (LGBM). Grid search optimization is applied to enhance the classifier’s performance. The results are evaluated using accuracy, precision, recall and the F1 score. The study achieves a maximum accuracy of 96.55% with GB, while AdaBoost attains the lowest accuracy at 93.91%.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predictive Modeling of Key Performance Indicators for Greenhouse Gas Emission Reduction Using Machine Learning

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Claudia Diamantini,
  • Tarique Khan,
  • Alex Mircoli,
  • Domenico Potena

摘要

As demand for digital technologies grows, so does the impact of greenhouse gas (GHG) emissions. This paper explores innovative strategies to reduce carbon dioxide (CO2e) emissions in the product and technology industry, focusing on sustainable solutions to minimize environmental impact during product processes and transformations. The integration of artificial intelligence and Key Performance Indicators (KPIs) is significant within industrial sectors. This study proposes a machine learning-based approach to predict KPIs aimed at minimizing GHG emissions across product processes and transformations. A dataset obtained from the Environmental Protection Agency (EPA) portal was used; feature selection is performed using the Recursive Feature Elimination (RFE) method. We extract relevant KPIs from the dataset and mathematically represent these KPIs. We train six machine learning (ML) classifiers: Random Forest (RF), Decision Tree (DT), K-nearest neighbors (K-NN), Gradient Boosting (GB), Adaboost, and Light Gradient Boosting Machine (LGBM). Grid search optimization is applied to enhance the classifier’s performance. The results are evaluated using accuracy, precision, recall and the F1 score. The study achieves a maximum accuracy of 96.55% with GB, while AdaBoost attains the lowest accuracy at 93.91%.