Modeling and forecasting of milk production in the western zone of Tamil Nadu Shankar S. Vishnu1,*, Ajaykumar R.2, Ananthakrishnan S.3, Aravinthkumar A.4, Harishankar K.5, Sakthiselvi T.6, Navinkumar C.7 1Department of Basic Sciences, Dr. Y.S. Parmar University of Horticulture and Forestry, Solan-173 230, Himachal Pradesh, India 2Department of Agronomy, Vanavarayar Institute of Agriculture, Pollachi-642 103, Tamil Nadu, India 3Department of Soil Science and Water Management, Dr. Y.S. Parmar University of Horticulture and Forestry, Solan-173 230, Himachal Pradesh, India 4Division of Plant Pathology, ICAR-Indian Agricultural Research Institute, New Delhi-110 012, India 5Department of Agricultural Economics, S. Thangapazham Agricultural College, Tenkasi-627 758, Tamil Nadu, India 6Department of Soil Science, Kerala Agricultural University, Vellayani-695 522, Kerala, India 7Department of Agricultural Meteorology, Vanavarayar Institute of Agriculture, Pollachi-642 103, Tamil Nadu, India *Corresponding Author: S. Vishnu Shankar, Department of Basic Sciences, Dr. Y.S. Parmar University of Horticulture and Forestry, Solan-173 230, Himachal Pradesh, India, Email: s.vishnushankar55@gmail.com
Online published on 1 November, 2023. Abstract Background In India, the dairy business is expanding dramatically. Tamil Nadu milk cooperatives significantly contribute to the growth of the dairy sector in the state. In terms of delivering economic income for dairy smallholders and satisfying customer demand, the identification of milk production is one of the primary financial operations made in India. Considering this, it is crucial to understand future production to enhance and sustain the sector's growth and development. Methods The present investigation attempts to predict and forecast milk production in Tamil Nadu using time series models. Yearly milk data from 1976 to 2020 was taken. The study considered Auto-Regressive Integrated Moving Average (ARIMA) and Artificial Neural Network (ANN) to select the appropriate stochastic model for forecasting milk production in Tamil Nadu. Further statistical modeling procedures employed for milk production reveal that the selection of a suitable time series model will always depend on the nature of the data. Result Results revealed that the ARIMA model is selected as the best model despite ANN, even if it is considered the most powerful model. The CAGR for forecasted milk production from 2020-2025 was 0.02%. Model adequacy criteria like RMSE, MAPE and MAE are used. Based on observation ARIMA model (1, 1, 2) is chosen as the best model over the ANN model. Top Keywords ANN, ARIMA, Milk Production, Time series models. Top |