스마트 농업의 작물 생존 예측을 위한 인공지능 분석 = (Artificial Intelligence Analysis for Crop Survival Prediction in Smart Agriculture)

다국어 초록 (Multilingual Abstract)

Smart agriculture has been evolving by integrating ICT, IoT, and AI technologies to maximize agricultural productivity and optimize resource utilization. This study aims to predict crop survival in smart agriculture using various machine learning and deep learning models while analyzing and comparing their performance. To achieve this, Random Forest, XGBoost, LightGBM, LSTM, and GRU models were implemented, and their predictive performance was evaluated using accuracy, precision, recall, and F1-score. SHAP analysis was applied to enhance model interpretability and assess the impact of key variables on prediction outcomes. The experimental results indicate that XGBoost and LightGBM demonstrate the highest predictive performance, confirming the effectiveness of tree-based boosting models in crop survival prediction. Notably, SHAP analysis reveals that variables such as pesticide usage type, estimated insect count, and pesticide application frequency significantly influence the prediction results. This study highlights the potential of AI-based predictive models in smart agriculture and emphasizes the importance of optimizing controllable environmental factors to improve crop survival rates.