<p>This paper investigates the problem of predicting the success of contracts concluded in Russia. It is based on a&#xa0;machine learning algorithm: gradient boosting over decision trees. The classifier parameters are adjusted, and the most important features are generated and searched for. The following important attributes were found: percentage of contract price drop; contract price per day; contract price per employee; contract price multiplied by price change.</p>

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

DETERMINING FULFILLMENT RATE OF PUBLIC PROCUREMENT CONTRACTS

  • M. K. Khromov,
  • A. V. Shokurov

摘要

This paper investigates the problem of predicting the success of contracts concluded in Russia. It is based on a machine learning algorithm: gradient boosting over decision trees. The classifier parameters are adjusted, and the most important features are generated and searched for. The following important attributes were found: percentage of contract price drop; contract price per day; contract price per employee; contract price multiplied by price change.