K-MEANS CLUSTERING ALGORITHM BASED CLASSIFICATION OF SOIL FERTILITY IN NORTH WEST NIGERIA

Soil fertility determines a plant's development process that guarantees food sufficiency and the security of lives and properties through bumper harvests. The fertility of soil varies according to regions, thereby determining the type of crops to be planted. However, there is no repository or any source of information about the fertility of the soil in any region in Nigeria especially the Northwest of the country. The only available information is soil samples with their attributes which gives little or no information to the average farmer. This has affected crop yield in all the regions, more particularly the Northwest region, thus resulting in lower food production. Therefore, this study is aimed at classifying soil data based on their fertility in the Northwest region of Nigeria using R programming. Data were obtained from the department of soil science from Ahmadu Bello University, Zaria. The data contain 400 soil samples containing 13 attributes. The relationship between soil attributes was observed based on the data. K-means clustering algorithm was employed in analyzing soil fertility clusters. Four clusters were identified with cluster 1 having the highest fertility, followed by 2 and the fertility decreases with an increasing number of clusters. The identification of the most fertile clusters will guide farmers on where best to concentrate on when planting their crops in order to improve productivity and crop yield.