A BIG DATA MODEL FOR AN INFERENCE ENGINE OF A PRODUCTION COMPANY

Recent trends in data processing and analytic systems prove that the amount of data generated in production companies are massive and handling them is a herculean task. Many data centers of such companies are currently facing efficiency dearth due to the increasing demand in the data analytics and processing. Cloud- based data centers are now adapted to the processing technology developed by platforms like Hadoop. Hadoop has provided a solution for high scale data processing with Map-Reduce Algorithm as the root. Data analytics is considered a vital aspect of Big Data handling. There are multiple models of Hadoop analytics available. In this paper, a multi- model analytics approach is proposed for the performance improvement of Hadoop. The model improves the dimension by performing a procedure of categorization of data and identification of missing values. With the help of this technical paper put forwards hybrid model of Hadoop analytics that is a combination of Hadoop with K-Mean Clustering Analytics and Hadoop with EM algorithm Analytics according to the data results of preprocessing. The model gives some faster analytics for Hadoop processed data which can result in improvement of an analytical system in a production company