Linear Attribute Projection and Performance Assessment for Signifying the Absenteeism at Work using Machine Learning

• DOI: 10.35940/ijrte.c4405.098319
• Publication Date: 2019-10-09T06:47:10Z
• Authors: M. Shyamala Devi*; Usha Vudatha; ; Sukriti Mukherjee; ; Bhavya Reddy Dont...; ; S B Adhiyan; ; Nallareddy Jishnu

In recent times, with the technological advancement industry and organization are transforming all their inflow outflow operations into digital identity. At outset, name of is also in hands employee. One major needs employee working environment to avail leave or vacation based on family circumstances. Based health condition need employee, must extend for satisfaction The performance predicted days organization. With this view, paper attempts analyze number hours by using machine learning algorithms. Absenteeism at work dataset from UCI Repository used prediction analysis. absent achieved three ways. Firstly, correlation between each attributes found depicted as a histogram. Secondly, top most high correlated features identified which directly fitted regression models like Linear regression, SRD RANSAC Ridge Huber ARD Regression, Passive Aggressive Regression Theilson Regression. Thirdly, Performance analysis done analyzing metrics Mean Squared Error, Absolute R2 Score, Explained Variance Score Log Error. implementation python Anaconda Spyder Navigator Integrated Development Environment. Experimental Result shows that have effective minimum MSE 0.04, MAE 0.16, EVS 0.03, MSLE 0.32 reasonable 0.89.