Stat2 Building Models For A World Of Data Pdf Download [2021]
Stat2 Building Models For A World Of Data Pdf Download https://urluss.com/2tanat
Rey and Wells [75] highlighted the challenge of forecasting with Big Data. The challenge arising from the large size of the data to be analysed cannot be overstated. The sheer volume of data recorded is a result of the fact that the growth in the volumes of data that are being collected and the exponential increase in the sizes of data that are generated by ubiquitous technologies (e.g., [5, 21, 24, 51, 57, 58, 89]) are creating an entirely new wealth of information. In fact, according to [89], an average of five gigabytes of data is created every second, with this figure expected to increase to more than 100 gigabytes per second in a few years. There are several reasons for this growth in data. For example, the Internet was the reason behind the exponential growth in the data in early years, with the advent of smartphones causing the growth of data in the coming years. According to [89], the growth of data is due to increase in the use of smartphones, the development of advanced technologies (e.g., [47], wearables, cloud computing, social media, and the Internet of Things among others) and the increase in the growth of mobile data. Companies such as Google, IBM, Microsoft, Facebook and Amazon generate vast amounts of data every day. In fact, Google generates about two petabytes of data every day (i.e., [87]) whilst Facebook produces over five petabytes of data per day (i.e., [84]).
The information presented in this paper is of direct relevance to the Journal of Forecasting and that it should be of interest to researchers in the field. A brief description of the topic is presented in section 1.1. In section 1.2, a detailed review of the relevant literature is presented. Section 1.3 presents the review of forecasting techniques utilized for forecasting with Big Data, whilst section 1.4 presents the review of applications. In section 1.5, we present a summary of the findings and the concluding remarks of this paper. Finally, we present the research methodology in section 1.6.
To build a model, the first step is to acquire an appropriate dataset for model development. However, [3] points out that the availability of Big Data is the driving force in the development of new methodologies for developing models, and thus there are a wide variety of Big Data-based methods from which to choose. This dependency on Big Data raises the issue of access to Big Data. [78] points out that Big Data are generated in a way that is not controlled by the user and therefore difficult to obtain. Additionally, [3] shares that Big Data is a commodity which is owned and controlled by the data provider. This has led to the emergence of the so-called ‘Data Mining Wars’ [2], a conflict between the data mining industry, and the data mining distributors. Whilst many users of Big Data believe that the problem of access to Big Data is created by the data miner, the reality is that the difficulty of access is created by the data miner himself when he does not supply the data. [3] points out that the issue of access to Big Data is not new, and was seen in the past, with the issue of access to the data being the driving force behind the development of new statistical techniques. 827ec27edc