Pl/sql design to determine the training data pattern on the Adaline neural network

Abstract

Inventories of goods in business activities greatly affect the smooth running of business activities. The durian fruit sales business is also affected by inventory. Durian fruit sellers must be able to predict the needs of durian fruit customers so that sales business activities run stable. For sellers with low durian fruit sales, they will have no difficulty predicting demand for durian fruit. But for sellers of durian fruit in large quantities, it will take a long time and is prone to errors. An error in the prediction causes the sales business to lose. It takes high accuracy and fast time to make predictions. Adaline neural network has been proven to be able to solve prediction problems precisely. Adaline network performance is affected by training data. Durian fruit sales data is used as network training data to improve network performance. The focus of this research is the design of the pl/sql database to determine the training data input pattern in the android application to predict the number of durian fruit customer needs. Types of pl/sql used in this study are procedures and functions. Meanwhile, the method used to design pl/sql is an extreme method.