Census Employee Salary Prediction using Supervised
Machine Learning
Journal:
GRENZE International Journal of Engineering and Technology
Authors:
Raghawendra Naik, Pavan N.Kunchur
Volume:
6
Issue:
2
Grenze ID:
01.GIJET.6.2.503_2
Pages:
214-219
Abstract
Payment plans are a key tactical area for fulfillment and growth of knowledge
based industry and also optimum salary offer is essential to retain high performance
employees. One of the challenges that industries face fairly often is finding such income
facts, based on several information about a current employee or a future employee. Given
the characteristics of a current employee or a future employee like his / her demographic
profile alongside other information like performance level, qualification, etc., prediction of
the salary class are often done by using many well-known machine learning algorithms. But
unluckily, those details of employee of any industry are generally not presented publicly for
performance evaluation of machine learning algorithms. In this paper, this limitation is
overcome to some extent by employing a public database (UCI census data set) which has
most of the attributes available for a segment of population for salary prediction. i.e., this
paper aimed at examining and investigating three well-known supervised machine learning
classifiers namely Gaussian Naive Bayes, KNN(K-Nearest Neighbors) and Decision Tree
Classifier using the UCI census data set to find out the best classification algorithm out of
above stated three well-known classifiers. It also aimed to determine the most effective
classifier to be used in this area. Finally from the investigation we found that KNN(KNearest
Neighbors) Classifier performed well in comparison with the opposite two
classifiers.