Even if you do not have a social network, it may be found Researchers with the University of Heidelberg anonymous Facebookcooperation, the network of friends as a test data set basis. Processing data through the Air Jordan 21 network analysis and machine learning methods, under Air Jordan 7 certain conditions in order to accurately predict whether acquaintance between Nike Air Max 90 VT 40% non-users. The study also shows that social networking potential of many non-users. Our friend who is a social network user, but I do not use social networks people can not recognize it? Researchers at Heidelberg University Interdisciplinary Scientific Computing Research Center, which issues investigated. As can be seen from their research work, through network analysis and machine learning methods associated with Mode association between users and Nike Air Max 95 between users and non-users were re-processed to obtain the relationship between the non-users. Using this simple correlation data, it can, under certain Nike Basketball Shoes conditions, be able to predict with 40% probability the two know each other whether non-social network users. Any social networking platform social groups will be divided Nike Dunk SB Mens into two parts, the user (black villain) and non-users, non-users and sub-associated non-users (red villain) and unrelated non-user (gray villain). Non-user e-mail contact with the red line marked, represents both knowledge and the gray line indicates an association was not observed, namely Non-users do not know each other. The relationship between non-users can and contact mode (green line) between the user and non-user to accurately deduce from the relationship between users (black line). In recent years, scientists have been working on this problem research, through the use of adequate learning and prediction algorithm to calculate the input data analysis conclude? In a social network, some users will not Jordan Spizike disclose the information out, such as sexual orientation or political affiliation, but if his friends provide enough information about him, then the result will be calculated higher accuracy. Heidelberg Image Processing Laboratory Cooperation (HCI) co-founder, Professor Hanpulache (Prof. Dr. Fred Hamprecht) said, once it has been confirmed that friendships are informed, then the machine learning, the prediction of some unknown content no longer is a big challenge. So far, these studies were limited to users of social networks, namely those with user Nike Air Max 89 files and consent (social network) to a person given privacy policy. \u0026 Ldquo; however, there is no such non-user privacy agreement, therefore, we are automatically generated so-called shadow file (shadow profiles) a study. \u0026 Rdquo; Professor Zweig in Scientific Computing Heidelberg University Interdisciplinary Research Centre (IWR) work (Prof. Dr. Katharina Zweig) explains. In a social network, suggesting that non-user information is possible, for example, by using so-called discovery applications friend. When a new one Facebook user registration, they are asked to provide a complete e-mail contact list, even those not Facebook users. \u0026 Ldquo; in a social network, such as who and who may know the user information will be in addition to the social network and know who the information Nike Air Max 90 Current tied up. In turn, this association may be used to infer the relationship between a considerable number of non-users \u0026 rdquo ;, Zweig colleague Professor Horvath (\u0026 Aacute; gnes Horv \u0026 aacute; t) says. Researchers at the University of Heidelberg, standards-based network analysis of the structure of machine learning program to complete the calculation. Because the Institute is not easily able to get data, researchers with Facebook anonymity cooperation, the network of friends as a test data set basis. Use a wide range of models to simulate possible to 2015 Latest Nike Shoes distinguish between users and non-users, such a distinction is used to verify the correctness of the results. The researchers used a standardized computer in just a few days you can figure out which non-users are most likely to be other people's friends. Scientists at the University of Heidelberg so surprised that all analog methods have had the same qualitative results. According to the Heidelberg Image Processing Laboratory Dr. Han Siman cooperation (Dr. Michael Hanselmann) argument, based on the ratio of actual assume social network users in the population accounted for, and they will be uploaded to the probability-mail address book online, the results can make We predict the relationship between the 40% of non-users, which represents compared with simple guess, the accuracy rate increased by 20 times. The results illustrate the potential of social networks many non-users. Professor Hanpulache stressed that the study only as a relational data base, which can not help but surprise. Many social networking platform, with more user information, such as age, income, education, experience, or the address and the like. Using these data, and then equipped with the appropriate technology infrastructure and other network analysis of structural characteristics, the researchers believe, forecast correctness will be greatly enhanced. Professor Zweig said: \u0026 ldquo; will have to say, our project illustrates that we, as a member of a social group, to give those who do not understand the relationship between the data provided may be used to what extent. \u0026 Rdquo; In this paper, the author Dai Jin i horse finishing translation from Sciencedaily shell network