Answers are grounded in this paper and cite their sources.
Soil moisture classification requires AI technology to avoid manual human inspection. The use of AI is also to realize smart farming, where agricultural activities can be more precise in classifying soil moisture. Several AI approaches can be used for data classification, particularly image data. One frequently used approach for classification is traditional machine learning. The classification stage requires feature extraction from the dataset as input. Non-parallel feature extraction is time-consuming, where the amount of data to be extracted is quite large. Therefore, this research proposes the use of parallel processing. Parallel processing has been shown to accelerate feature extraction by more than twice compared to non-parallel methods. The resulting feature extraction vectors are used as input for supervised learning. This research compared the use of a single classifier with stacking. Accuracy was improved between the use of a single classifier and stacking. The accuracies achieved using SVM, Random Forest, and Logistic Regression were 0.9157, 0.9105, and 0.8621, respectively. Meanwhile, stacking increased accuracy to 0.9275, a 1.2% increase over SVM. This research also compared the deep learning approach using the MobileNetV2 model. Accuracy was also superior using the stacking approach. These results are expected to provide insights into how parallel processing can accelerate data processing. Furthermore, the combination of supervised learning and data classification accuracy can be improved.
No takes yet. Share an insight, caveat, or question.
Adhinata et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: