Time Series Model to Forecast the Pricing of Dairy Products

Abstract

Seasonal patterns and patterns of food production are linked with each other, which contributes to have a significant impact on the economy of a country. Seasons and food production patterns are linked with each other. Because of this, it is of the utmost importance to establish projections about the production of food patterns that are sensitive to variations in the climate, which will ultimately result in satisfied customers and successful production. Because of this, it is vital to establish trustworthy techniques of dairy production forecasting in order to prevent a shortage of dairy production in an industry. As a result, the purpose of this study is to construct a model that is capable of reliably predicting the price of dairy products in relation to changes in the climate. In the course of the research for the thesis, Ireland will serve as the case study, and a mix of time series and machine learning models will be utilized in order to make price predictions. A time series of data relating to dairy production from the year 1990 to 2019 is extracted and utilized as a data source for dairy products. This data source will be used until the year 2019. The work that was actually done for the proposed paper makes use of the entire dataset for training purposes and makes use of the pricing list for the previous year for the testing phase. These are then utilized as variables in order to assess product yields as well as product losses. The implementation, on the other hand, is based on the principles of time series as well as four machine learning algorithms, specifically ARIMA, ARIMA Garch, SEM, and SARIMA. In compared to the results of other models, it was shown that the SARIMA model produced superior results. In addition, the findings were computed on the basis of an assessment matrix that took into account the root mean square error.