دانلود رایگان مقاله لاتین آنالیز داده بزرگ از سایت الزویر
عنوان فارسی مقاله:
تکنیک های موازی برای تجزیه و تحلیل داده های بزرگ در نسخه جدید سرویس ارزیابی معاملات آتی
عنوان انگلیسی مقاله:
Parallel Techniques for Large Data Analysis in the New Version of a Futures Trading Evaluation Service ☆
سال انتشار : 2015
برای دانلود رایگان مقاله آنالیز داده بزرگ اینجا کلیک نمایید.
مقدمه انگلیسی مقاله:
1. Introduction
With the fast development of economy, more and more investors put money into futures markets with the expectation of making profit. A futures market has the characteristics of highrisk and high-return due to its high leverage ratio. A large number of non-professional investors participate in futures trading. Some of them make profits, and most of them lose money because of their different capabilities. For these participants, what they are most concerned about is to receive some recommendations from experts according to their past trading transactions. We design a futures trading evaluation system to help them. The system is a profit making capability assessment software, which is developed by the authors and GT Futures Brokerage Company. The system processes historical transaction data of a specific investor, and extracts key performance indicators and trading behavior patterns. Through the analysis of these indicators and pat-terns, with the help of the knowledge of investment experts, the system gives customers some suggestions on trading capabilities, trading habits, and trading psychology, etc. to help customers identify their shortcoming and improve themselves in the future. In short, according to a customer’s historical transaction data, the system generates an assessment report, with the hope that it can help the customer improve his (her) trading capabilities, stop loss, and make profit. The futures trading evaluation system tries to discover some valuable information, i.e., the customer’s behavior characteristics buried in the process of trading, from large volume of historical transaction data. In the evaluation system that we built, the volume of the data is greater than 200 GB, and it is growing at a pace of 300 to 500 MB per day. The data should be analyzed as timely as possible, so that customers can adjust their trading strategies according to the analytic results. Our contributions in this paper include: (1) we propose a segmented bit encoding of dimensional tables for star schema data, which can eliminate join operation during aggregating on any level of the hierarchy of any dimension; (2) we exploit inter customer data scan sharing, which can improve report generation throughput greatly; (3) we have conducted a series of experiments to evaluate the proposed techniques.
برای دانلود رایگان مقاله آنالیز داده بزرگ اینجا کلیک نمایید.
کلمات کلیدی:
Tools and techniques for massive data analysis | SCAI www.hpc.cineca.it › Training › List Events The purpose of this workshop is to present scientists tools and techniques, ... from the scientific community for its applicability in large parallel data analysis. Parallel Techniques for Large Data Analysis in a Futures Trading ... doi.ieeecomputersociety.org/10.1109/GCC.2010.45 Parallel Techniques for Large Data Analysis in a. Futures Trading Evaluation Service System. Qin Xiongpai1,2, Wang Huiju1,2, Du Xiaoyong1,2, Wang Shan1,2. Design and Analysis of Large Data Processing Techniques https://pdfs.semanticscholar.org/b23c/a1dc6f70dfe97101c3665d44061b0a237c53.pdf by M Vaidya - 2014 - Cited by 1 - Related articles top of DFS. Large-scale data analysis is run in parallel on data stored in DFS. ... technique which is adopted by Google adopting as a major technology for graph ... Techniques and Environments for Big Data Analysis - Springer www.springer.com/cn/book/9783319275185 Presents the recent advances in the fields of Big Data Analysis, as well as the ... learning can play a vital role in big data analysis by adopting Parallel, Grid, and ... Applied Parallel Computing. Advanced Scientific Computing: 6th ... https://books.google.com/books?isbn=354048051X Juha Fagerholm, Juha Haataja, Jari Järvinen - 2003 - Computers Parallel execution of different data mining algorithms and techniques can be ... can be parallelized to make them effective in the analysis of very large data sets. Large-Scale Data Analytics - Page 42 - Google Books Result https://books.google.com/books?isbn=1461492424 Aris Gkoulalas-Divanis, Abderrahim Labbi - 2014 - Computers MapReduce offers a simple interface to implement parallel data analysis tasks. ... criticize the lack of schema information, sophisticated data access methods ... Handbook of Big Data Technologies - Page 69 - Google Books Result https://books.google.com/books?isbn=331949340X Albert Y. Zomaya, Sherif Sakr - 2017 - Computers Many big data analysis applications work on batch-wise input, and there are many like ... techniques like data parallel, task parallel or graph parallel techniques.
