Identifikasi shorea menggunakan k-nearest neighbour berdasarkan karakteristik morfologi daun
Abstract
Dipterocarpaceae is a group of tropical plants that is used in a timber industry. One of Dipterocarpaceae clans is Shorea, that is the best timber-producing plant in the industrial world. Shorea is difficult to be identified because it has a lot of diversity. The inability to recognize Shorea in forest will enlarge the exploitation of Shorea that has a good timber quality, and silviculture work becomes less of the target because it is not known which Shorea species that will extinct. Shorea tree is usually identified by using the stems, leaves, fruits, and flowers. However, leave is used for the identification in this research because it tends to be available as a source of observation at anytime. The leaves in this reasearch are the collection from Bogor Botanical Gardens. Data are obtained by manual calculating to get the characteristics of the leaves. The obtained data will be processed using k-Nearest Neighbor to get the closeness of new data and training data. The leaves that are included in this research are Shorea multiflora, palembanica, balangeran, lepida and assamica, each one has 10 data. Each set of data has 10 attributes that support the leaf characteristics. From the whole data, Shorea are divided into five subsets including data training and data testing for each subset. This research has two experiments the first experiment without normalization produces 84% accuracy and the second experiment with a normalization produces 100% accuracy.
Collections
- UT - Computer Science [2323]