آی اس آی پیپر

دانلود رایکان مقاله انگلیسی ISI با موضوع استخراج کلان داده ها با رایانش موازی

عنوان فارسی مقاله:

استخراج کلان داده ها با رایانش موازی: مقایسه روش های توزیعی و MapReduce (نگاشت-کاهش)

عنوان انگلیسی مقاله:

Big Data Mining with Parallel Computing: A Comparison of Distributed and MapReduce Methodologies

دانلود رایگان مقاله ISI با فرمت PDF:

مشاهده توضیحات کامل و خرید ترجمه فارسی با فرمت ورد تایپ شده:

بخشی از مقاله انگلیسی :

2. Literature Review

2.1 Distributed Data Mining 

Distributed computing can refer to the use of distributed systems to solve computational problems. In particular, a problem is divided into many tasks, each of which is solved by one or more computers (or processors) that run concurrently in parallel. In addition, each processor can communicate with each other by message passing (Coulouris et al., 2011). In the traditional data mining approach, the data are usually centralized and a specific algorithm is then chosen to process and analyze the data under a single computing platform. However, for a big data problem or large scale data mining, this is not so simple, and the performing the data mining tasks under the distributed computing platform has become an important area of research investigation (Zaki, 2000; Zheng et al., 2012). Generally speaking, the objective of distributed data mining is to perform the data mining tasks based on the distributed resources, including the data, computers, and data mining algorithms (Park and Kargupta, 2002). Figure 1 shows a general distributed data mining framework where different data sources may be homogenous and/or heterogeneous. Each data mining algorithm handles its corresponding data source under a single computing platform leading to a local model. Then, these local models are aggregated in order to generate the final model.