Dimensional Reduction Method based on Big Data Techniques for Large Scale Data

Abstract

In the age of big data, analysing complex and enormous amounts of data costs time and money and might result in mistakes and misunderstandings. Therefore, bad inference and decision-making, and occasionally irreversible and catastrophic occurrences, could result from imprecise and mistaken reasoning. On the other hand, effective administration and use of priceless data can considerably expand knowledge and lower costs through preventive measures. Time-to-event and survival data analysis are the cornerstones of risk assessment in this area and play a crucial part in estimating the likelihood of numerous occurrences, including the failure of a device or component. As a result, before beginning any analytic process, it is desirable to apply appropriate approaches to effectively reduce large-scale, huge, and complex data, especially in terms of variables. To minimize the abovementioned decision-making challenges and to make survival data and failure analysis easier, we offer an applied data reduction strategy in this research that enables us to obtain appropriate multiclass classification in large dimensionality data and huge datasets. In order to estimate the risk in addition make decisions in the context of a complex analysis of big volume in survival analysis, this research presents an applied analysis data and attribute selection methodology.