Sugarcane variety identification using Dynamic Weighted Directed Acyclic Graph Similarity

Utomo, Adi Heru and Sarno, Riyanarto and Ginardi, R.V. Hari (2016) Sugarcane variety identification using Dynamic Weighted Directed Acyclic Graph Similarity. 2016 International Conference on Information & Communication Technology and Systems (ICTS). pp. 226-230. ISSN 2338-185X

	Text (Full text) utomo2016.pdf - Published Version Available under License Creative Commons Attribution Share Alike. Download (254kB)
	Text (Hasil Similarity) Hasil Turnitin - Sugarcane.pdf - Supplemental Material Available under License Creative Commons Attribution Share Alike. Download (1MB)
	Text (Hasil Peer Review) Sugarcane Variety Identification Using Dynamic Weighted Directed Acyclic Graph Similarity.pdf Download (98kB)

Abstract

Dynamic wDAG Similarity algorithm can be applied to sugarcane annotation. At first, we have to make a wDAG structure of many different varieties of sugarcane. We also have to make wDAG of sugarcane that will be annotated. Then, we have to calculate the similarity between wDAG types of sugarcane that will be annotated and wDAG of all the existing types of sugarcane. This similarity calculation results will present sequence similarities ranging from the most similar to the most distant from sugarcane varieties were annotated. This Dynamic wDAG Similarity algorithm has difference compared with the previous wDAG Similarity algorithm. WDAG used in this research has the node labeled , arc labeled and arc weighted, where the weight of the arc can be changed dynamically. This research fixes the previous studies of static wDAG, in which the weight values on the arc of wDAG can not be changed. On Dynamic wDAG, the weight on each arc is based on the fuzzy calculations that show the tendency of sugarcane varieties were annotated. And the fuzzy value is calculated based on agronomic traits of sugarcane to be annotated. Leaf node is the part of wDAG that will be compared first. The similarity calculation result between the two wDAG is affected by data on a leaf node to be compared and the weights of the arcs. The result shows that this method gained the average of Precision of 96%, the average of Recall of 88.5%, and the average of Accuracy of 96%.

Actions (login required)

View Item