Exporation And Research Of ODM Data Mining To Forest Management In Tahe

Now the application of various data mining technologies is becoming the hot issue in the academic field home and abroad. Application data mining in business field has gained perfect success. However, the successful application in forest production has rarely been seen. This thesis relies on the project of informational construction of digital forestry in Tahe, bases on the second-degree data base in Tahe, takes the decision of forest management as the mining objective and takes ODM provided by Oracle as the tool for data mining. Combined various data mining technologies with the scientific theory of forest management, this thesis clearly explains the procession of data mining of forestry, makes a comperison on the application of two important technologies of data mining——Classifier of Decision Tree and Classifier of Support Vector Machine. Through the comparison analysis of the exact results of the two classifier technologies, this theis testifies the feasibility and veracity of application the classifier of Decision Tree to forest. Meanwhile, through the optical processing and abstracting principle of generating decision tree, it constructs a principle collection of Malin forest farm in Tahe ,which is suitable for land and tree, and provides decision for scientific forest management in Tahe. The key points and research center of this thesis are the application of various data mining technologies to the procession of data mining of forestry. It is mainly consisted of the following parts:First, it introduces the task, theory, main procession and present situation of application of various data mining technologies in forestry.Second, it introduces the main theories in this data mining procession. It contains the theory of Principal Component Analysis, MDL, data discretization, decision tree and support vector machine.At last, using the data mining tool of ODM provided by Oracle, this thesis takes the data mining model of Malin farm in Tahe forest bureau as an example, and introduces deliberately the application of various data mining technologies in forestry data mining. Through the comparison analysis of the veracity of the two application results, testifies the feasibility of the technologies and the veracity of the results.