Skill-Centric Career Path Optimization through Machine Learning Models

View at publisher Summarise with VIRAT
Authors: Farooq Sunar Mahammad, Bokkasam Rithika, Kummari Bharathi, Badiginchula Lalitha, Gundam Sandya Rani, Merva Naga Mallikarjuna
Cited by: 0
References: 18

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Individual mentorship has been replaced by academic opportunities and, therefore, made machine learning-based employment recommendation systems more significant. It will be hard to find the right job and study the chosen profession without the old means of decision making that does not consider the individual qualities of each person. This was done through the creation of a synthetic dataset with the use of samples of Kaggle, which is a place where anyone could share datasets. The data contains the educational background of an individual, interests, abilities, and talents. A display of the results was done using Matplotlib and Seaborn. Linkages were made to clean the data through the use of a label encoder to make uniform and understandable results. We constructed several models of categorisation, including the Random Forest and SVM. Also, more accurate model was predicted with the assistance of XGBoost, and to make it even more reliable, a hybrid Voting Classifier utilizing XGBoost, CatBoost, LightGBM, and Random Forest was introduced. The data was divided into two, the first part was used to test the model, and the second part was used to train the model. We made use of Accuracy, F1-score, Recall, and Precision to evaluate the performance of the model. XGBoost was found to be the best of all the models that were tested with a score of 99.6% in all of the tests. The outcomes were relatively close to those of XGBoost, and Voting Classifier also turned out to be rather successful. To make the model predictions easier to understand, we applied LIM-based explainability, and to ensure that users would be able to quickly interact with the system and get its predictions immediately, we applied the Flask framework using SQLite. The prescriptions of the direction taken by the task are more precise, as it is much more effective in predicting anything in the whole system.

Fields of Study & Subjects

AI and HR Technologies Recommender Systems and Techniques Data Mining and Machine Learning Applications

Publication details

Journal / source
International Conference on Image Processing and Capsule Networks (ICIPCN)
Pages
1653–1658
Publication date
27 January 2026
Language
EN