Implementation of Conjugate Gradient Bacpropagation Algorithm in Predicting the Percentage of Formal Labor in Sumatran Pulau

Julius Sitanggang (STIKOM Tunas Bangsa Pematang Siantar, Indonesia)
Verdi Yasin (STMIK Jayakarta)
Solikhun Solikhun (STIKOM Tunas Bangsa Pematang Siantar, Indonesia)

INTERNATIONAL JOURNAL OF MECHANICAL COMPUTATIONAL AND MANUFACTURING RESEARCH
Vol. 11 No. 3 (2022): November: Mechanical Computational And Manufacturing Research

Formal workers work for companies as trained workers or those whose job status is employees and employees. They get stronger legal protection and official employment contracts and are in organizations that are legal entities. This paper aims to prove and implement the conjugate gradient backpropagation algorithm that can predict data on trends in formal labour on the island of Sumatra's impact Covid-19. The research data is on the percentage of formal workers on the island of Sumatra in 2015-2021, whose source is the Central Statistics Agency (BPS). This study uses 5 architectural models, namely 5-10-1, 5-15-1, 5-20-1, 5-25-1, 5-25-1, 5-30-1, 5-35-1. Of the five architectural models that have been used, the best architecture is obtained, namely the 5-20-1 model with a perf result of 0.000000002, literacy 292 in 4 seconds