Scholarlink

This work presents an AI system designed to assist in solving coding problems by recommending appropriate data structures. We developed a stacked ensemble model comprising hyperparameter tuned boosting models, focusing on array, graph, and string data structures to facilitate our tagging system. This system categorizes coding challenges from Codeforces and Leetcode, achieving notable effectiveness with a precision-recall area under the curve (PR-AUC) of 0.815 and an accuracy of 77.8%. Our approach employs grid search for hyperparameter optimization and stratified k-fold validation, ensuring robust model performance. The system supports programming education by enhancing understanding and efficiency in problem-solving, setting a foundation for future expansions to include more comprehensive programming concepts and refined performance metrics.

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

Stacking of Hyperparameter Tuned Models for Tagging Coding Problems

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Sathya Krishnan Suresh,
  • S. Lakshmana Pandian,
  • P. Shunmugapriya

摘要

This work presents an AI system designed to assist in solving coding problems by recommending appropriate data structures. We developed a stacked ensemble model comprising hyperparameter tuned boosting models, focusing on array, graph, and string data structures to facilitate our tagging system. This system categorizes coding challenges from Codeforces and Leetcode, achieving notable effectiveness with a precision-recall area under the curve (PR-AUC) of 0.815 and an accuracy of 77.8%. Our approach employs grid search for hyperparameter optimization and stratified k-fold validation, ensuring robust model performance. The system supports programming education by enhancing understanding and efficiency in problem-solving, setting a foundation for future expansions to include more comprehensive programming concepts and refined performance metrics.