Crop Production Prediction Model Using Fuzzy Associative Memory Based on Agroclimatic and Land Characteristic Variables

Model Prediksi Produksi Tanaman Menggunakan Fuzzy Associative Memory Berdasarkan Variabel Agroklimat dan Karakteristik Lahan

Abstrak

Crop production prediction is important to support agricultural planning, resource allocation, and early decision-making in farming systems. Crop yield is influenced by various factors, including land characteristics and agroclimatic conditions. This study aims to develop a crop production prediction model based on agroclimatic and land variables using the Fuzzy Associative Memory (FAM) method. The study used secondary data from a crop production dataset containing land area, average air temperature, rainfall, and production variables. The modelling stages included determining input and output variables, constructing fuzzy sets, defining triangular membership functions, developing FAM rules, forming F AM matrices, performing model examination, and conducting defuzzification using the winner-takes-all approach. The input variables consisted of land area, air temperature, and rainfall, while the output variable was crop production classified into small, medium, and large categories. The results show that FAM can map linguistic input conditions into production categories through structured fuzzy rules. This model can be used as an initial approach for crop production prediction, especially when dealing with limited data and uncertain agroclimatic conditions.