Data Mining Management

Questions: write about the following requirements. 1. What: What is the problem? What are the requirements needed in order to solve that problem? 2. Why: Why do you need to solve that problem? 3. How: How is it to be solved? Please focus on an optimal solution. 4. What are the lessons learn from this research paper? 5. Please also provide: Summary, Conclusion and Recommendations Answers: The problem; and the requirements needed in order to solve that problem According to the research paper this is clearly seen that the data mining is basically the extraction of knowledge. But during that there is some of the problem that involved with the data mining. These are basically three types of data mining problems these are the clustering: together group similar items and to dissimilar ones separate. The next one is the the values of predict to attributes someone from the other training data. Analysis association: detect the condition of attribute-value that to occur the frequently together (Blockeel et al., 2011). The requirement needed to solve clustering problem is:(1) cluster manager, (2) file system shared by clustered, (3) administration of DB2 cluster. (4) Physical hosts, (5) members of DB2, (6) facilities of cluster caching (Aeron, Kumar Moorthy, 2012). Reason for solve that problem To solve this problem this is very important, that the analysis of cluster is the grouping set task of the objects in the such way that to the same group object (this is called cluster) are more similar (in some kinds of sense or the another) to each other than those to in other type of groups (clusters). This is the main task of data mining exploratory, and the statistical data analysis common techniques. These are basically used in many fields. Therefore, clustering can be formulated as the optimization of multi-objective problem. At the time of data mining, the groups of resulting are the interest matter, in the classification automatic the discriminative resulting power is of interest. That is often to leads misunderstandings between the coming researches from the data mining fields and the learning of machine, since to the use of same terms and often the algorithm same, but in different goals (Lv, 2015). Solution of the problem There are many way to solved this problem these are Hierarchical Methods:These are the method that the cluster constructs by partitioning recursively the instances in either the bottom-up or top-down fashion. The methods that are sub divided as the following manner. Clustering of hierarchical agglomerative- Each of the object represents initially own its cluster. These are the clusters merged successively until the desired cluster structure is obtained. The clustering hierarchical Divisive - In general every object are belonging initially from the cluster one. Then after that clusters are divided into the sub-clusters, that to divide into successively own sub-clusters. Clustering of the single link- That the methods to the distance consider between the clusters that must be shortest distance equal to from any of the one cluster member to any of the of the member to another cluster. Link of Complete Clustering - The methods that to the distance considering in between the clusters two to be equal to the longer from distance any of the member of cluster one to any of the member of another cluster. Clustering Average-link These methods that basically consisting distance between the clusters two that equal to be the distance from average any of the members of cluster one to the any of the member of other cluster (Saha, 2012). Methods of Partitioning: This method is also very important in this context because from this method the instances relocate by them moving from one to another cluster, that starting from the partitioning initial. These kinds of methods are requiring typically for the cluster number will be users pre-set. To, globally optimally achieved in the clustering of partitioned-based, these enumeration exhaustive process of the all the required partitions possible (SajjatulIslam Zainal Abedin, 2013). Algorithms of error minimizing: The algorithms, that to tend well with work compact and clusters isolated, these are the frequently most and used intuitive methods. The ideas of basic that are to clustering find that structure to minimize the creation of certain errors that to measured the distance of the every instance that is to value representative (Zheng, 2014). Clustering of graph-theoretic: This is the methods that produced via graphs clusters. The graph edges are connected represented as instances nodes. The theoretic well known algorithm graph is basically based on the MST. Density-based methods: In this method to the points assume to belong the cluster each drawn from distribution of probability specific. The distribution overall of data is to be assumed the distribution of mixtures several. Methods of model based clustering: This is the method that to attempts for optimizes the fit between the data given and some of the models of mathematics. Methods of Grid-based: These are the methods that are space partition into the finite of cells number that to form a structure grid on which the options for performed clustering (Zeng Xiao, 2014). Lessons learn from this research paper The lessons that are learn from this paper are the problem of data mining and how to solve this problem, the reasons for solve this problem, and the solution of this problem. These are the lessons learn from this paper (SajjatulIslam Zainal Abedin, 2013). Conclusion and Recommendations At the end of this study, data mining is basically the extraction of knowledge. But during that there is the problem that involved with the data mining. These are basically the clustering problem. To solve this problem this is very important, that the analysis of cluster is the grouping set task of the objects in the such way that to the same group object (this is called cluster) are more similar (in some kinds of sense or the another) to each other than those to in other type of groups (clusters). This is the main task of data mining exploratory, and the statistical data analysis common techniques. At the time of problem solving there is some of the methods these are the Methods of Partitioning, Methods of Hierarchical, methods of Density-based, methods based on clustering of Model, and Methods of the based- Grid. Reference List Aeron, H., Kumar, A., Moorthy, J. (2012). Data mining framework for customer lifetime value-based segmentation.Journal Of Database Marketing Customer Strategy Management,19(1), 17-30. doi:10.1057/dbm.2012.1 Blockeel, H., Calders, T., Fromont, ., Goethals, B., Prado, A., Robardet, C. (2011). An inductive database system based on virtual mining views.Data Mining And Knowledge Discovery,24(1), 247-287. doi:10.1007/s10618-011-0229-7 Lv, K. (2015). Study on Pharmaceutical Database Management Based on Data Mining Technology.J. Inf. Comput. Sci.,12(8), 2979-2986. doi:10.12733/jics20105831 Saha, S. (2012). Application of Data Mining in Protein Sequence Classification.IJDMS,4(5), 103-118. doi:10.5121/ijdms.2012.4508 SajjatulIslam, M., Zainal Abedin, M. (2013). Impacts of Data Mining on Relational Database Management System Centric Business Environments.International Journal Of Computer Applications,75(3), 21-27. doi:10.5120/13091-0371 SajjatulIslam, M., Zainal Abedin, M. (2013). Impacts of Data Mining on Relational Database Management System Centric Business Environments.International Journal Of Computer Applications,75(3), 21-27. doi:10.5120/13091-0371 Zeng, J., Xiao, Z. (2014). Automatic Mining and Processing Dormancy Data in the Database Management System for Small and Medium Enterprises.AMM,513-517, 1927-1930. doi:10.4028/www.scientific.net/amm.513-517.1927 Zheng, R. (2014). Simulation of Data Mining System Design in Database.AMR,989-994, 2020-2023. doi:10.4028/www.scientific.net/amr.989-994.2020