Detection of changes in the employment environment in Japan based on the numbers of people leaving and entering employment using NMF
- 10.2991/jrnal.2017.3.4.11How to use a DOI?
- resilience, data mining, machine learning
If there were no changes in the environment surrounding businesses, the numbers of people leaving and entering employment would stay almost the same. Therefore, understanding the numbers allow us to make assumptions about the changes inside and outside companies. However, when categorizing businesses into industry sectors and clusters of business, you will see that the numbers of people leaving and entering employment have been nearly opposed for the last 15 years, and it is difficult to detect changes in the employment environment of Japan’s businesses. This study tried to improve the sensitivity of detecting changes by applying NMF (non-negative matrix factorization) into the Survey of Employment Trends. While businesses maintain the number of people they employ at a certain level because of severe restrictions, we assumed they respond to the surroundings by changing the composition of employment. Accordingly, we identified the correlation between the numbers of people leaving and entering employment in each sector characterized by employment patterns that we found by applying NMF. As a result we successfully improved the sensitivity level of detecting changes, which we would like to report in this study.
- © 2013, the Authors. Published by Atlantis Press.
- Open Access
- This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).
Cite this article
TY - JOUR AU - Masao Kubo AU - Hiroshi Sato AU - Akihiro Yamaguchi AU - Yuji Aruka PY - 2017 DA - 2017/03/01 TI - Detection of changes in the employment environment in Japan based on the numbers of people leaving and entering employment using NMF JO - Journal of Robotics, Networking and Artificial Life SP - 265 EP - 269 VL - 3 IS - 4 SN - 2352-6386 UR - https://doi.org/10.2991/jrnal.2017.3.4.11 DO - 10.2991/jrnal.2017.3.4.11 ID - Kubo2017 ER -