Scholarlink

This paper presents an ensemble learning-based framework for forecasting food wastage into this scenario, managing variability in data, dependency on certain environmental conditions, and socio-economic differences. For carrying out feature selection technique RFE is applied using a wide dataset that considered environmental conditions, socio-economic indicators, and historical records on such phenomena. For evaluation purposes, various ensemble models are selected: random forest, gradient boosting, LightGBM, and so forth and evaluated using MSE, RMSE, R \(^2\) , and MAE as metrics. The technique gave very accurate predictions and withstood dropouts, and the feature’s importance analysis provided valuable information to the stakeholders/policymakers on how to design better sustainable food distribution strategies.

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

Enhanced Prediction of Food Wastage Using Ensemble Learning and Feature Selection

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Venkata Kamya Punugupati,
  • Bhargavi Maridu,
  • Vijayalakshmi Maddiboyina,
  • Akshay Kanuri,
  • Ruthvik Peddineni

摘要

This paper presents an ensemble learning-based framework for forecasting food wastage into this scenario, managing variability in data, dependency on certain environmental conditions, and socio-economic differences. For carrying out feature selection technique RFE is applied using a wide dataset that considered environmental conditions, socio-economic indicators, and historical records on such phenomena. For evaluation purposes, various ensemble models are selected: random forest, gradient boosting, LightGBM, and so forth and evaluated using MSE, RMSE, R \(^2\) , and MAE as metrics. The technique gave very accurate predictions and withstood dropouts, and the feature’s importance analysis provided valuable information to the stakeholders/policymakers on how to design better sustainable food distribution strategies.