Clementine12 in all of the following data mining algorithms is Professor Xie Bangchang Data Mining (Data Mining) dozens of analytical methods, so that everyone preliminary understanding of the model, however, is also frequently encountered in everyday mining algorithms, we hope to useful! (Even a data mining company, with one of these algorithms can be unrivaled in the world) a memory-based reasoning (Memory-Based Reasoning; MBR) main memory-based reasoning is a concept (case) best buy hours to a known case predict future cases of some of the attributes (attribute), usually find the most similar cases for comparison. Memory based reasoning method best buy hours has two main elements, namely a function of distance (distance function) binding function (combination function). Distance function intended to identify the most similar case in; combined function will combine the properties of similar cases, for forecasting purposes. Advantage of memory-based reasoning is that it is susceptible to various types of information, best buy hours which does not need to obey certain assumptions. Another advantage is its ability to learn, it can be an old case study to get knowledge about the new case. The criticism is higher than it needs a lot of historical data, there is sufficient historical data can only make good predictions. Besides memory-based reasoning method in handling also more time-consuming and difficult to find the best combination of distance function and function. Detection, customer reaction may range applications include fraud prediction, best buy hours medical diagnosis, classification and other aspects of the reaction. 2, market basket analysis (Market Basket Analysis) The main purpose of market basket analysis is to identify best buy hours what kind of things should be put together? Commercial applications in by customer buying behavior to understand best buy hours what these customers and why customers buy these products, find the relevant association (association) rules, companies get these rules by mining interests and build competitive advantage. For example, the retail utility allows analysis of changes in product design or shelving arranged best buy hours on commercial packages to attract customers and so on. The basic operation of the market basket analysis process consists best buy hours of the following three points: (1) select the correct items: correctly referred to here but for enterprise best buy hours body, it must be in the hundreds of thousands of items, best buy hours select a really useful The items out. (2) to explore co-occurrence matrix via (co-occurrence matrix) of excavated association rules. (3) to overcome the practical limitations: the more the selected items, resources, best buy hours and time-consuming calculation of the longer (exponential increment), then you must use some techniques to reduce the loss of resources and time. Basket analysis techniques can be applied on the following questions: (1) For credit card purchases, customers may be able to predict what the future purchase. (2) For the purposes of the telecommunications and financial services, through market basket analysis to design best buy hours different services portfolio to maximize profits. (3) Insurance can be detected basket analysis may not insure unusual combination and as a precaution. (4) For patients on the combination regimen, whether the market basket analysis can be used as a combination of these medications can cause complications judgment basis. 3, the decision tree (Decision Trees) tree has a strong ability in solving classification and prediction, it is a way to express the law and these rules places a series of questions that come out, continue to ask questions via the ultimately exported desired results. The top is a typical tree roots at the bottom of many leaves, it will record best buy hours broken down into different subsets, each subset field may contain a simple rule. Furthermore, the decision best buy hours tree may have different shapes, such as binary trees, ternary tree or a mixture of tree types. 4, genetic algorithms (Genetic Algorithm) cellular genetic algorithms to learn the process of evolution, cells can continue through the choice of reproduction, crossover, mutation produces new cells better. Mode of operation is very similar to the genetic algorithm, it must be a good pre-established pattern, and then through a series of similar operations generate new cellular processes, to determine whether the offspring best buy hours produced by this model agree with the use of appropriate function (fitness function), only the last The most consistent results can survive, this program has been operating until this function to converge to the optimal solution. Genetic best buy hours algorithms on a cluster (cluster) problem has a good performance, generally available to assist memory-based reasoning and neural network applications. 5, cluster detection techniques (Cluster Detection) This technology covers best buy hours a wide range, including genetic algorithms, neural networks, statistical analysis of the cluster have this feature. Its goal is to identify groups of similar data previously unknown in many analyzes, are beginning to use the cluster detection technology best buy hours to serve as the beginning of the study. 6, Link Analysis (Link Analysis) link analysis is based on the mathematical graph theory (graph theory), based on records by the relationship between the development of a model, which is based on the relationship of the body, by the people, things and the development of the relationship between people and objects, or objects of a considerable number of applications. best buy hours For example, by means of telecommunications services to customers using link analysis to collect phone time and frequency, and then infer why customer preferences, in favor of the company's proposed plan. In addition to the telecommunications industry, best buy hours more and more marketing industry best buy hours Zheyi do favor the use of link analysis research enterprise. 7, online analytical processing (On-Line Analytic Processing; OLAP) Strictly speaking, online best buy hours analytical processing and data mining is not particularly a technology, but through online analytical best buy hours processing tools, consumers can more clearly understand the information hidden potential implications. Like some visual processing technology in general, by way of a chart or graph to show, in general, the feeling best buy hours will be more friendly. Such tools can also assist to transform the data into the target information. 8, neural networks (Neural Networks) neural network based on repetitive learning methods and learning will pay a bunch of examples, it is sufficient to conclude a distinguished style. If faced with a new example of the neural network to learn in accordance with its past results are summarized, exported to push new results are part of a learning machine. best buy hours Issues related best buy hours to data mining can also be taken to learn the way the nervous type, which can make learning very correct prediction function. 9, discriminant analysis (Discriminant Analysis) encountered a problem when it's the dependent variable is qualitative (categorical), and the argument (predictor best buy hours variables) quantitative (metric) when, discriminant analysis is a very appropriate technology, typically used in solving classification The above problems. Ruoyin variables posed by the two groups, called the dual group - discriminant analysis (Two-Group Discriminant Analysis); If constituted by a number of groups, it is called multiple discriminant analysis best buy hours (Multiple Discriminant Analysis; MDA). (A) identify a linear best buy hours combination of predictor variables, so the ratio between the two groups with respect to the variation of the maximum variation in the group, and each a linear best buy hours combination of the linear best buy hours combination has been obtained previously not associated. (2) test whether the center of gravity of each group are different. (3) to find out which variables have the greatest difference between the forecast ability. (4) According to the forecast values best buy hours of the variables of new subjects, the subjects assigned to a group. 10 logistic regression analysis (Logistic best buy hours Analysis) When the difference between the analysis groups do not meet the normal distribution assumption, logistic regression analysis is a good alternative. Logistic regression analysis does not predict events (event) has occurred, but predicting the probability of the event. It argument assumes that the relationship between the dependent variable is the shape of the S-line, when the argument is very small, the probability value is close to zero; When the argument values slowly increase the probability value increases along the curve, to a certain extent when the curve Association rate begins to decrease, so the probability of a value between 0 and 1.
2013 Today in History: Naruto 649 words: Ninja Will 2013: Death 551 words: The Burnt Offerings 2013: One Piece 723 words: Change best buy hours Combat 2012: View sql server 2000 blocking lock and handle 2012: Using T-SQL statement to achieve Information such as SQL server and XLS file exchange 2012: MERGE in 2008 2012 SQL Server: sqlplus command line administrator login 2012: Data Mining: How Companies Now Know Everything About You 2012: DB2 Maintenance Manual 2012: CCB OCRM System 2012: [ DB2] db2pd tool 2012: Perl download page 2012: SPSS papers or reliability of the questionnaire analysis 2012: SPSS Shortcuts 2012: spss common operations Digest 2012: SPSS function Daquan: Arithmetic Functions 2012: SPSS function best buy hours Daquan: string functions 2012: [ SPSS Frequencies Process 2012: [SPSS] Recode, Compute, Count, If command 2012: six kinds of mining Weapons 2012: spss utilization best buy hours program 2012: Database Marketing and Data Mining 2012: Data Mining and CRM 2012: Data Mining Common Terminology 2012: data Experience 2012 mining process: data mining analysis tools 2012: Customer Segmentation 2012: time management principles 40 2012: data analysis to onlookers 2012: Implementation of the project's 2012 CRM rational thinking: What is CRM (customer relationship best buy hours management) 2012: Effective CRM 2012: Analytical CRM 2012: Customer Information Management and Analysis 2012: Data Mining Applications for CRM 2012: In CRM, the effective use of data mining 2012: [CRM] BLENDS Model 2012: customer segmentation technology or art? 2012: Failure to connect to Oracle databases and solutions introduced 2012: Connecting heterogeneous best buy hours data sources in Oracle Database best buy hours 2012: Oracle large text access to solve the problem in the ASP 2012: VB write scan tool 2012: [VB6.0] merge multiple a Word file 2012: make contacts with VB 2012: VB trigonometric Hot Links Web Design (16) Ajax (82) facebook (3) ASP (50) ASP.NET (555) FACEBOOK grammar (8) Apache (89) Tomcat (96) HTML (21) IIS (75) SilverLigh (362) JSP (163) CSS Teaching (115) Java Script (201) PHP Teaching (189) mobile web (10) pages of new information (2)
Android (33) Mindwave Mobile (0) NFC category (2) gestures class (gesture) best buy hours (2) Drivers category (2) Transmission best buy hours Class (transfer) (2) for the log (Log) (29) Application category (11) drawing class (Draw) (86) Foundation Classes (41) Advanced Class (343) Sensors (Sensing) (32) photography class (Take Picture) best buy hours (94) system class (System) (575) APK class (40) Q & A Class (139) XML class (10) Collection class (44) Tools (Tools) best buy hours (199) interface class (Interface) (392) Map class (137) Internet class (Network) (243) Site Category (Website) (29 ) Communication (77) Hardware category (Hardware) best buy hours (215) Safety class (70) Class Bluetooth (Bluetooth) (88) games category (Game) (54) Multimedia (Multimedia) (274) simulator class (Emulator ) (110) Database class (Databases) (103) SD memory cards (46) Free Software (Software) (108) the latest information (Information) (73)
iPhone (21) Website category (7) drawing class (4) photography best buy hours class (3) Touch class (4) ios (5) ipad (4) games category (10) map class (16) Safety class (5) Communication (18) Advanced (83) Tools (45) Hardware category (33) interface (78) Safety class (12) system class (159) Internet class (93) Multimedia (89) Database class ( 22) IPHONE (109) the latest information (20)
Business courses (21) Product Management (7) e-commerce (31) body language (2) Carnegie (7) Crisis Management (24) BI (4) ERP (38) CRM (66) OLAP (3) SPSS (22 ) Data Mining (42) Market Research (1) Information Retrieval (4) Management (17) Knowledge Management (14) Human Resources (31) Accounting & Finance (28) the stock of knowledge (13) Investment (22) Data analysis (1 ) Logistics Marketing (15)
No comments:
Post a Comment