Besides offline and online information analysis, big data has a number of defining characteristics: volume, velocity, variety, and veracity. Added to these characteristics are a number of general features relevant to data science, such as validity, volatility, variability, viscosity, value, and visualization. The importance of big data does not start from quantity but from its use. Big data helps organizations create new growth opportunities. High-performance solutions based on algorithms and methods of high accuracy and computing power are aimed at processing huge amounts of ready-to-collect data and quickly tracing trends and correlations.
Structured in eight chapters, this book presents key big data concepts through case studies and examples, giving readers the benefit of learning from real marine problems and real marine big data. Large data flows with detailed spatial and temporal measurements, as well as the size and processing of outsourced data, are topics of great interest for researchers and practitioners who want better knowledge of the marine environment; information technology (IT) specialists who approach these topics through solutions based on big data will also benefit. The methods and techniques for monitoring and manipulating marine data based on machine learning and deep learning, as well as the methods for evaluating and including them in modern knowledge bases, are new challenges in the field.
The book starts off by covering basic topics, for example, marine data types based on historical and current marine observations from various world sources. Aided by a diverse range of methods and techniques for the acquisition and classification of marine big data, readers will benefit from the support provided by theoretical research and modern technologies related to analysis, mining, quality control, and security.
Identifying, analyzing, and evaluating risks as well as making appropriate decisions in the event of disasters are taken into account in the process of adopting solutions and technologies for processing marine big data information. Major system changes due to incidents or after the application of risk controls are also considered. Chapter 5, “Application of Marine Big Data in the Shanghai Storm Surge Disaster Assistant Decision-Making System,” is dedicated to an application framework based on the Spark parallel architecture system, Hadoop framework, MapReduce programming model, and 3D visualization technologies.
Chapter 6, “Monitoring of Ocean Oil Spills Based on Marine Remote Sensing Big Data,” shows how remote sensing techniques can be used to determine the degree of pollution in a marine environment. The chapter also looks at the development of solutions based on cloud computing and the processing technologies of massive remote sensing images. The methods and applications used to manage and analyze remote sensing images can be used to monitor oil spills.
In chapter 7, “The Application of Marine Big Data to Sea Ice Classification with Deep Learning,” the authors propose a deep learning architecture for sea ice classification based on residual networks (ResNets) and synthetic aperture radar (SAR) images. The presented deep learning methods use big data from images obtained via active sensors (the SAR method) or passive sensors (passive microwave sensors and altimeter and thermal infrared sensors).
Big data is based on a series of technologies that are constantly evolving. Big data analysis in the cloud, predictive analysis, time series analysis, and deep learning are some of the technologies addressed here. This book is ideal for researchers and practitioners new to marine big data; some experience with the basics of big data is assumed.
