Preface

Ann. Data. Sci. (2014) 1(2):149–150 DOI 10.1007/s40745-014-0019-3 Preface Theory, Methods and Applications in Data Science Yingjie Tian Published online: 28 October 2014 © Springer-Verlag Berlin Heidelberg 2014 This issue of 2014, Annals of Data Science (Volume 1, No. 2) presents 7 papers from the several areas of Data Science. They are contributed from 20 authors and the coauthors come from 5 countries and regions: Chile, China, Iran, Serbia and USA. These contributed papers deal with data science problems from three aspects: the first is the theoretical base of data science, the second is about the methods of data mining, and the third contains the applications. For the theoretical base of Data Science, the paper “Factor space, the theoretical base of data science”, by Pei-Zhuang Wang, Zeng-Liang Liu, Yong Shi and Si-Cong Guo, introduced factor space theory, which provides a general coordinate system to describe the real world and a theoretical base for data science. Factor space was published in the same year coincidently with the formal conceptual analysis and rough sets. The three branches were the pioneers in intelligence mathematics, but the former one had focused on genetic analysis for uncertainty several years. It is a bridge connecting randomness and certainty and also a bridge connecting fuzziness and certainty. Based on the theory, factorial databases is presented, which carries a new kind of statistics to do intelligent analysis for coming tide of Big Data. For the methods of Data Science, the paper “Review on: Twin Support Vector Machine”, by Yingjie Tian and Zhiquan Qi, closely reviewed Twin Support vector machines (TWSVMs) and provided an insightful understanding of current developments. As the useful extension of the traditional SVM, TWSVM has lower computational complexity and better generalization ability, therefore in the last few years it has been studied extensively and developed rapidly, and became the current researching Y. Tian (B) Research Center on Fictitious Economy & Data Science, Chinese Academy of Sciences and Key Laboratory of Big Data Mining and Knowledge Management, Chinese Academy of Sciences, Beijing 100190, China e-mail: 123 150 Ann. Data. Sci. (2014) 1(2):149–150 hot spot in machine learning and data mining. This paper also pointed out their limitations and highlighted the major opportunities and challenges, as well as potential important research directions. The paper “Extended Exponential Geometric proportional hazard model”, by Sadegh Rezaei, Sina Hashami and Lotfollah Najjar, proposed the Extended Exponential Geometric (EEG) proportional hazard model. Researchers often use ordinary least square and generalized linear models even for censored data, while some researchers presented a useful method for cases which include censored data and used this model without considering baseline hazard models. These methods are all described and compared with the EEG proportional hazard model, the simulation results show that this model provided more accurate predictions of mean, median, and high cost cases and this model can replace to exponential hazard models and OLS with and without log translation and semi-parametric proportional hazard. Milan Stanojevic, Bogdana Stanojevic, and Nina Turajli proposed a discrete multipleobjective linear fractional programming (MOLFP) model for the web service selection problem in the paper “Optimization of multiple-objective web service selection using fractional programming”. Due to the fact that a large number of available services offer similar functionality, when choosing actual services to be included in the composition their non-functional properties must also be taken into account. On the other hand certain constraints regarding the required performances may also be given. Therefore, web service selection presents a multiple-objective multiple constraint problem and can be modeled as the MOLFP. They presented a complete methodology for solving this problem and reported the experimental results. Applications of Data Science include three papers, the paper “Ranking Countries by Medal Priorities Won in the 2014 Sochi Winter Olympics”, by Thomas L. Saaty, Xiaoyue Liu and Michael Sanserino, used Analytic Hierarchy Process (AHP) to quantify the priorities of different games according to environmental and people factors and also quantify the priorities of gold, silver and bronze medals, then use these priorities to compute the total scores of all three types of medals won by each country in order to determine the ranking of the countries which won medals in the 22st Winter Olympics held in Russia. The paper “A Business Model Design for the Strategic and Operational Knowledge Management of a Port Community”, by Felisa Córdova and Claudia Durán, proposed a business model design for managing a sea port community, to transmit the knowledge that have been created and acquired so that all the agents participating in the community can share this business knowledge. The paper “Commercial Banks with A Hybrid Prediction Model”, by Yinhua Li, Yong Shi, Anqiang Huang and Haizhen Yang, proposed a new prediction model combined by trait recognition and SVM, which used the accounting data measured one year prior to identity the features of problem banks. The new method outperformed nine popular prediction models in overall accuracy. It was also shown that ROA, liquidity assets and short-term gaps are sound predictors for bank failure prediction. As Volume 1, No 1 of ADS has attracted many readers, we believe that this Issue will still keep the effect. ADS encourages the contributors around the world to address different challenging problems in Data Science, including the theories, methods, applications, especially the corresponding problems arising in Big Data environment. 123  (...truncated)