Comparison of data mining classification techniques : a case study in forest cover type prediction | |
| Author | Aye Mon Tun |
| Call Number | AIT RSPR no.CS-15-03 |
| Subject(s) | Data mining Forest mapping--Remote sensing |
| Note | A research submitted in partial fulfillment of the requirements for the degree of Master of Science in Computer Science, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Research studies project report ; no. CS-15-03 |
| Abstract | Data mining techniques have been used in many fields for effective discovery and prediction of new knowledge, like vegetation prediction. Classification is one of such predictive data mining techniques and used to predict unknown class labels of new data based on the trained data of known class labels. In forest management, forest inventory plays a vital role and support as a core repository of forest data. Availability of accurate forest data for forest inventory is often a very time-consuming, labor-intensive and costly task. Using data mining classification techniques can help the task. This research examines three commonly used data mining classification techniques:Support Vector Machines (SVM), Random Forest (RF)and k-Nearest Neighbors (k-NN),and compares their classification abilities of correct forest cover type by using forest cover type data consisting of 15120 training data records and 565892 test data records, each with 54 attributes.The research study develops accurate predictive models for SVM, RF and k-NN by means of feature selection and parameter tuning.According to the results, three classifiers gains high level of accuracy using different number of features and parameters and SVM with the highest accuracy is the best of three classifiers for forest cover type prediction. |
| Year | 2015 |
| Corresponding Series Added Entry | Asian Institute of Technology. Research studies project report ; no. CS-15-03 |
| Type | Research Study Project Report (RSPR) |
| School | School of Engineering and Technology (SET) |
| Department | Department of Information and Communications Technologies (DICT) |
| Academic Program/FoS | Computer Science (CS) |
| Chairperson(s) | Guha, Sumanta; |
| Examination Committee(s) | Vatcharaporn Esichaikul;Dailey, Matthew N.; |
| Scholarship Donor(s) | Ministry of Foreign Affairs, Norway; |
| Degree | Research report (M. Sc.) - Asian Institute of Technology, 2015 |