SN Computer Science Computer Graphics and Computer-Aided Design

Supervised and Unsupervised Machine Learning Methods Applied for Enrolling Women in the Canadian Armed Forces

Calitoiu, Dragos, Ueno, Ryuichi, Boyd, Peter
Springer Nature · 2022-04-26 · DOI 10.1007/s42979-022-01115-y

Abstract

A tailored military recruiting approach that focuses on a subset of population with highest recruiting potential is being considered by the Canadian Armed Forces. In this paper, a logistic regression model (supervised machine learning) was developed using 4 years of historical applicant data. The score obtained was used to rank Canadian postal codes and to identify the ones with the highest potential for recruitment of women. Additional filtering was applied using marketing segments provided by a vendor. The final selection was clustered (unsupervised machine learning) based on the collective social media behaviour of each postal code and was binned using the distance to the nearest recruiting centre. The clustering and the binning were derived to select the optimum marketing channel and message.

Read at publisher →