<![CDATA[The freshness of milkfish is influenced by several factors, one of which is the time of storage of fish. In the process of storage and processing at traders and households, it is still done manually and causes errors in determining the freshness of milkfish. To overcome these problems, a tool that can determine the freshness of milk fish will be designed quickly and automatically. In making this tool an arduino microcontroller and MQ135 gas sensor will be used to detect ammonia, and TCS3200 sensor to detect the RGB color of milkfish. The results of the two sensors in the form of 4 parameters or features will be used to determine the freshness of milkfish with the Naive Bayes method ... The Naive Bayes method was chosen because this method is very flexible if there are changes to the training data, and requires little training data to can do Naive Bayes calculations, and finally the results of the classification method are also quite accurate. From the testing carried out starting from the sensor testing method and computational time the result is the TCS3200 error percentage when detecting RGB meat is 2.2%. In testing the sensor MQ135 sensor correlation value obtained with an output voltage of 99.22%. For testing methods using 100 training data and 18 test data, classification using Naive Bayes obtained an accuracy of 94.4% with an average computing time of 2.7 seconds.]]>
Copyrights © 2019
