In this work we determine the optimal epsilon value on peatland on DBSCAN Algorithm to clustering data on peatland hotspots in sumatera. DBSCAN is a base algorithm for density based data clustering which contain noise and outliers. We found using this method that the area which has the highest density of hotspots in Sumatra in 2013 peatland is contained in cluster 1 of Riau Province that is equal to 2112 hotspots.

Indonesia has experienced extensive land-cover change and frequent vegetation and land fires in the past few decades. We combined a new land-cover dataset with satellite data on the timing and location of fires to make the first detailed assessment of the association of fire with specific land-cover transitions in Riau, Sumatra. During 1990 to 2017, secondary peat swamp forest declined in area from 40,000 to 10,000 km² and plantations (including oil palm) increased from around 10,000 to 40,000 km². The dominant land use transitions were secondary peat swamp forest converting directly to plantation, or first to shrub and then to plantation. During 2001–2017, we find that the frequency of fire is greatest in regions that change land-cover, with the greatest frequency in regions that transition from secondary peat swamp forest to shrub or plantation (0.15 km⁻² yr⁻¹). Areas that did not change land cover exhibit lower fire frequency, with shrub (0.06 km⁻² yr⁻¹) exhibiting a frequency of fire >60 times the frequency of fire in primary forest. Our analysis demonstrates that in Riau, fire is closely connected to land-cover change, and that the majority of fire is associated with the transition of secondary forest to shrub and plantation. Reducing the frequency of fire in Riau will require enhanced protection of secondary forests and restoration of shrub to natural forest.

Tree counting is an important plantation practice for biological asset inventories, etc. The application of precision agriculture in counting oil palm trees can be implemented by detecting oil palm trees from aerial imagery. This research uses the deep learning approach using YOLOv3, YOLOv4, and YOLOv5m in detecting oil palm trees. The dataset consists of drone images of an oil palm plantation acquired using a Fixed Wing VTOL drone with a resolution of 5cm/pixel, covering an area of 730 ha labeled with an oil palm class of 56,614 labels. The test dataset covers an area of 180 ha with flat and hilly conditions with sparse, dense, and overlapping canopy and oil palm trees intersecting with other vegetations. Model testing using images from 24 regions, each of which covering 12 ha with up to 1000 trees (for a total of 17,343 oil palm trees), yielded F1-scores of 97.28%, 97.74%, and 94.94%, with an average detection time of 43 s, 45 s, and 21 s for models trained with YOLOv3, YOLOv4, and YOLOv5m, respectively. This result shows that the method is sufficiently accurate and efficient in detecting oil palm trees and has the potential to be implemented in commercial applications for plantation companies.

A Fuzzy Based Decision Support System for Evaluating Land Suitability and Selecting Crops

Hotspot is one of forest and land fires indicator that is used for developing fire early warning system. This study aimed to determine the areas that have high density of hotspot in the peat land area of Sumatra. The method applied was Density-Based Spatial Clustering Algorithm with Noise (DBSCAN) to form hotspot clusters in Sumatra in the years of 2002 and 2013. This study found that there were changes in the pattern of distribution of hotspot in peat land from 2002 to 2013, where in 2002 were mostly found in moderate depth peat land, whereas in 2013, mostly found in very deep peat land area

(1) Background: Feature selection is the biggest challenge in feature-rich sentiment analysis to select the best (relevant) feature set, offer information about the relationships between features (informative), and be noise-free from high-dimensional datasets to improve classifier performance. This study aims to propose a binary version of a metaheuristic optimization algorithm based on Swarm Intelligence, namely the Salp Swarm Algorithm (SSA), as feature selection in sentiment analysis. (2) Methods: Significant feature subsets were selected using the SSA. Transfer functions with various types of the form S-TF, V-TF, X-TF, U-TF, Z-TF, and the new type V-TF with a simpler mathematical formula are used as a binary version approach to enable search agents to move in the search space. The stages of the study include data pre-processing, feature selection using SSA-TF and other conventional feature selection methods, modelling using K-Nearest Neighbor (KNN), Support Vector Machine, and Naïve Bayes, and model evaluation. (3) Results: The results showed an increase of 31.55% to the best accuracy of 80.95% for the KNN model using SSA-based New V-TF. (4) Conclusions: We have found that SSA-New V3-TF is a feature selection method with the highest accuracy and less runtime compared to other algorithms in sentiment analysis.