OPTIMIZING FRUIT CULTIVATION THROUGH AN UNSUPERVISED LEARNING APPROACH WITH DATA SCALING TECHNIQUES FOR STRATEGIC INSIGHTS
This study employs K-Means clustering to analyze durian farm yield and area of production in Eastern Thailand, revealing patterns that could guide agricultural improvements. By employing advanced data preprocessing techniques including Min-Max scaling and Z-Score normalization, the research ensures robust, equitable consideration of all variables, enhancing the analytical process. Particularly, Min-Max scaling proved most effective for this dataset, optimizing the influence of each variable in the clustering process. The clustering categorizes farms by yield and land characteristics and uncovers optimal practices for durian cultivation. This approach offers actionable insights that promise to support sustainable farming practices and contribute to economic development in rural area. Determining that five was the optimal number of clusters was critical in identifying the most distinct and relevant patterns within the data. The findings highlight the potential of machine learning in agriculture, advocating for a data-driven strategy to optimize durian farm output and sustainability
Сведения
- Опубликовано
- 2025 · Т. 13, № 3, С. 112-127
- Язык
- Английский
- ISSN
- 2306-6172 · 2308-9822
- Издатель
- L. N. Gumilyov Eurasian National University