A Model of Online Market Analysis using Machine Learning in Big Data

Abstract

newline Advent of technology has given immense data in every sector such as business, science, web newlineapplication and manufacturing. Data can be of any type, it can be structured, semi-structured newlineor un-structured. Huge size of Online data comes under the category of Big data and it is newlineessential to train data using machine learning newlineMachine learning can be defined as computational methods using the experience to improve newlinethe performance or to make accurate predictions. Machine Learning is evolved from the study newlineof pattern recognition and computation learning theory in Artificial Intelligence, Machine newlineLearning explores the study and construction of algorithms over come following strictly static newlineprogram instructions by making data driven predictions or decisions by building a model newlinefrom sample inputs. To solve any machine learning problem there are 4 steps that need to be newlinefollowed. Collect data for one. Secondly pick the Model. Third train the model and then newlinefinally test the Model. newline Big data is uncountable and is so huge. That it can go beyond the size of hundreds of newlineTerabytes. Bigdata are structured as well as unstructured data. It has its root in all the fields newlinelike science, engineering, Technology and healthcare. In every second the size of Bigdata is newlineincreasing rapidly. Traditional database system has no capacity to handle such massive newlinequantities of data. The world is creating 2.8 quintillions of data per day from unstructured newlinedata sources like social media platform like Facebook, Instagram, mail, hike, photos, files etc. newlineBigdata forms the three level structures of data produced e.g structured data, newlineunstructureddata, semi structured data. newlineOur proposed algorithm New Modified K-Means Clustering Algorithm is based on S K newlineAhammad Fahad et.al.(2016), work A Modified K-Means Algorithm . This algorithm is newlineused in online market analysis model. We have taken more than one lakhs data from online newlinemarket such as Flipkart, Amazon and Snapdeal. Online market is dealing with numbers of newlinedata daily. Lot of