Single Node Hadoop Cluster for Small Scale Industrial Automation

Abstract

Data analytics is a key requirement for the growth of any industry or organization. It is required to solve various consumer and product predictions and insights in a way to benefit the organization as well as enhance profits. However, today’s small-scale industries with low budgets and huge data often face relinquished profit issues, which can be solved by a combination of big data warehouse systems and efficient data analytics tools. The paper takes into account the problems faced due to lack of data analytics integrated to big database structures, efficient data mining and analytics tools in the absence of effective databases. In this paper, we look on a developed working idea for embedding Hadoop (Apache Hadoop is an open-source framework built on top of JAVA which allows distributed processing with large datasets) with predictive and statistical data analytic tools like the Artificial, Convolutional, Generative neural networks and machine learning algorithms like Gradient boosting (XGBoost), support vector machine (SVM) and regularization regression models. We have considered the erroneous noise existing in data and used multilevel neural networks to solve this problem. We have made use of blob segmentation and detection for better identification of hotter regions.