Scholarlink

<p>The placement of students through campus recruitment is a major concern for students, institutions, and hiring companies. Many existing classification techniques focus solely on predicting whether a candidate will secure a job or not. However, it’s important to recognize that students often make multiple attempts before achieving success, and the probability of placement increases with each attempt. Conventional classification methods often overlook the influence of the number of attempts required to secure a job. We develop a new model that generalizes Logistic Regression to better forecast the likelihood of securing employment through campus placements incorporating the number of attempts. The model also finds the significance of other influential factors, such as academic performance, skill training in the placement process. Additionally, we propose a new model to predict the number of attempts needed to secure placement when a student is predicted as placed. Both models are evaluated on real campus placement data and they demonstrate superior performance compared to existing methods. We have also examined the impact of training on the likelihood of securing a placement.</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

Generalization of Logistic Regression to Improve Prediction: An Application on Training and Placement Data

  • 期刊论文

  • 发表日期：

    2025年9月1日

查看全文
  • Md Jakir Hossain Molla,
  • Hare Krishna Maity,
  • Biman Chakraborty,
  • Sk Md Obaidullah,
  • Soumya Sen

摘要

The placement of students through campus recruitment is a major concern for students, institutions, and hiring companies. Many existing classification techniques focus solely on predicting whether a candidate will secure a job or not. However, it’s important to recognize that students often make multiple attempts before achieving success, and the probability of placement increases with each attempt. Conventional classification methods often overlook the influence of the number of attempts required to secure a job. We develop a new model that generalizes Logistic Regression to better forecast the likelihood of securing employment through campus placements incorporating the number of attempts. The model also finds the significance of other influential factors, such as academic performance, skill training in the placement process. Additionally, we propose a new model to predict the number of attempts needed to secure placement when a student is predicted as placed. Both models are evaluated on real campus placement data and they demonstrate superior performance compared to existing methods. We have also examined the impact of training on the likelihood of securing a placement.