![]() | Only 14 pages are availabe for public view |
Abstract The objective of this thesis is to examine the most effective data mining, classification algorithms to extract new knowledge and information from agriculture crops for discovering the prediction rules for rice crop diseases data set. By applying suchclassification algorithms, we can discover and as well as gain understanding of the reasons for crops diseases. The classification algorithms of data mining can be applied successfully to predict Egyptian crops diseases. The various classification algorithms (i.e., decision trees, neural networks, random forest, Bayesian networks and random trees) are applied toEgyptian crops diseases to predict the occurrence of diseases. The selected classification algorithms are compared using diverse performance evaluation methods The experimental results indicate that these approaches are applicable and efficient and the accuracy of such classification algorithms differs according to the learning methodology and classification rule used. Identifying the best classification algorithm among all available is also evaluated. It is seen that for a given dataset, most of used algorithms are found to outperform any other and their performances are comparable. However, the decision tree algorithm has the higher predictability and understandable rules. |