Application of hidden Markov model in yam yield forecasting

Providing the government and farmers with reliable and dependable information about crop yields before each growing season begins is the thrust of this research. A four-state stochastic model was formulated using the principle of Markov, each state of the model has three possible observations. The model is designed to make a forecast of yam yield in the next and subsequent growing seasons given the yam yield in the present growing season. The parameters of the model were estimated from the yam yield data of Niger state, Nigeria for the period of sixteen years (2001-2016). After which, the model was trained using Baum-Welch algorithm to attend maximum likelihood. A short time validity test conduct on the model showed good performance. Both the validity test and the future forecast shows prevalence of High yam yield, this attest to the reality on the ground, that Niger State is one of the largest producers of yam in Nigeria. The general performance of the model, showed that it is reliable therefore, the results from the model could serve as a guide to the yam farmers and the government to plan strategies for high yam production in the region.