Maize Yield Prediction using Interval-Valued Fuzzy Sets as a Tool for Attaining Food-Security in Kaduna State

Abstract

Crop yield prediction is one of the most important pillars that have the potential to increase productivity and improve food security. As Maize is the largest grain crop grown almost everywhere in Kaduna State, its accurate yield  estimation using environmental data and growth data before harvest is critical for food security and agricultural policy development. In this paper, a Maize Yield Prediction model using Interval-Valued Fuzzy Sets (IVFS) is developed, which allows each factor to be represented by an interval membership value rather than a single crisp value, to predict maize yields in the northern, central and southern zones of Kaduna State. It is discovered that the southern zone has a very high yield potential with an estimated yield of 6.1-7.5 tons/ha, followed by the central zone with 5.2 - 6.6 tons/ha, then the northern zone having 4.3-5.8 tons/ha. It is shown that the northern zone is Suitable for drought-tolerant maize varieties while the southern zone is the most ecologically suitable zone for maize production, thereby making it the best for large-scale maize commercialization among all the three zones. It is demonstrated that the state experienced a north-to-south increase in predicted maize interval, with the Northern Zone having the lowest predicted range and the Southern Zone having the highest. This pattern reflects differences in climate and other factors, including soil, fertilizer, hybrid seed, herbicide, and mechanization. The IVFS model effectively represents yield as an interval, capturing possible productivity ranges and supporting zone-specific agricultural planning and resource allocation.