Data Analytics Model for Manufacturing Industry

Abstract

Manufacturing Industry (MI) has problems with Value of Gross Output (VGO), Input Cost (IC), and Value Added (VA) in productivity, investment, trendline, and estimation. To overcome this problem, we carry out data analytic using descriptive model (K Means Clustering/KMC) for productivity, diagnostic model (Naïve Bayes Classifier/NBC) for investment, predictive model (Linear Regression/LR) for product trendlines, and prescriptive model (Monte Carlo Simulation/MCS) for input cost estimation. The results of KMC are 3 clusters. The results of NBC are VGO, VA, and IC influenced by number of establishments, workers engaged, and labor cost. The results of LR shows a trendline model. The results of MCS are 3 IC scenarios. We summarize that high productivity will open up new investment opportunities supported by a linear trend of value of gross output and value added with low input costs.