Early Appearance

Abstract
Since ancient times, meat has been a fundamental part of the human diet and continues to be so in many cultures. Despite variations in the amount and source of meat consumed across different countries and cultures, meat remains a primary component in most Western diets, often accompanied by vegetable side dishes. Additionally, meat is considered an important factor in Türkiye as part of gastronomic traditions, celebrations, and events. In addition to its biological content, the significant presence of sensory features in meat is among the reasons for consumer preference. Meat is an excellent food from a nutritional perspective, providing all essential amino acids and many vitamins (B vitamins, particularly B12) and minerals (such as zinc and iron). It supports muscle synthesis and maintenance in the body, which is important for both physical function and metabolic health. Meat also contains important biologically active compounds such as taurine, creatine, hydroxyproline, carnosine and anserine. Given the complex interplay of factors affecting meat consumption, this study aims to estimate and forecast meat consumption in Türkiye using machine learning algorithms. Data from 1990 to 2023, including Gross Domestic Product (GDP), meat production, meat prices, feed prices, agricultural GDP, population, imports, and exports, were analyzed using Random Forest, Gradient Boosting, Support Vector Machine, AdaBoost, Neural Network, and Linear Regression models. The results indicate that Gradient Boosting and AdaBoost algorithms provided the most accurate predictions, highlighting the importance of agricultural GDP, meat production, and population in forecasting meat consumption.
Keywords: Gastronomic tradition, meat consumption, machine learning, prediction modeling, demand forecasting