CHICKEN MEAT SALES FORECAST BY USING MACHINE LEARNING TECHNIQUES A CASE STUDY OF SMALL ENTREPRENEURS IN LOM SAK DISTRICT PHETCHABUN PROVINCE

Abstract

           This research presents broiler sales forecasting using Machine Learning Techniques. A case study of small entrepreneurs in Lom Sak District, Phetchabun Province. The objectives are 1. To develop a Machine Learning Technique model suitable for forecasting broiler sales. Case study of small business operators in Lom Sak District Phetchabun Province 2. To compare forecasts with Machine Learning Techniques and Time Series Techniques. For forecasting broiler sales using Machine Learning Techniques. Case study of small business operators in Lom Sak District Phetchabun Province. By collecting data on broiler sales. From May 2023 to February 2024, which is secondary data obtained from small entrepreneurs of the broiler business in Lom Sak District. Phetchabun Province. Total for a period of 10 months and used to develop models using Machine Learning Techniques consists of 3 techniques: Neural Network Technique, Support Vector Machines Technique and Random Forest Technique. Develop the model using two Time Series Techniques, namely The Holt-Winters Exponential Smoothing Technique and The Auto Regressive Integrated Moving Average (ARIMA) Technique and then compare the performance of the models. The results of the research found that the best technique of Machine Learning Techniques is the Random Forest technique, which gives the lowest error rate. Time series is the ARIMA technique which gives the lowest forecast error rate.