Auxiliary-lag dependent gaussian process model for forecasting using proposed kernels and multi-start optimization method

Abstract Informed decision making is an important aspect based on improved and reliable forecasting. Modeling and forecasting complex real data are crucial for policymaking. In this paper, we have presented an auxiliary-lag dependent Gaussian process, a Bayesian non-parametric machine learning model, for improved forecasting using auxiliary lags. We also introduced some new multi-featured kernel functions that are versatile in dealing with seasonal and non-seasonal data. For comparison of the proposed model, the autoregressive random forest model, autoregressive artificial neural network model, seasonal autoregressive moving average models, and exponential smoothing models were used for forecasting rainfall data, climate modeling, and sugarcane production data, agriculture crop modeling, to show the versatility of the proposed models. Results confirmed the superiority of the proposed model over conventional models. The proposed methodology will be helpful for other researchers and local experts in making more reliable forecasting, which will be helpful in policymaking relevant to agriculture systems, water management systems, climate change, and natural disasters such as droughts and floods.