<![CDATA[Black Tilapia or which has the scientific name Oreochromis Niloticus is a type of freshwater fish that is widely consumed throughout the world, especially in Indonesia. Tilapia production continues to increase along with the increasing demands to meet food needs both domestically and abroad. One of the efforts made to achieve this target is cultivating tilapia seeds to improve the quality and production of tilapia. In cultivating tilapia in order to obtain quality fish yields, one of the most important factors influencing this is the quality of the water in the pond where the tilapia is cultivated. In Indonesia there are still many tilapia cultivators who measure and control pond water quality manually or even don't do this at all. For this reason, a water quality classification system will be designed for black tilapia cultivation using the support vector machine method which can carry out classifications to determine water quality in ponds automatically. The system is implemented using 3 sensors, namely the temperature sensor DS18B20, the pH sensor PH-4502c, and the turbidity sensor SEN0189. In the test, 3 features were used in the system to make training and testing data, namely temperature features, pH features, and turbidity features to determine the class of pond water being tested. 60 total training data consisting of 30 pieces of data for the "good" class and 30 pieces of data for the "bad" class were used to conduct training data for the classification system. The support vector machine classification method that is programmed in the system will carry out the classification process by reading the values ​​from the sensors and storing them as features to be processed and compared with the results of the training data. The system will look for the y value as the reference value for class classification results where if the y value ≤ 1 then the water will be declared "good" and if y> 1 then the water is said to be "bad". The test results are obtained by reading the output results on the system LCD for 15 seconds. From the test results, 15 reading were obtained where 12 of the test results were correct readings and 3 were incorrect readings. From the test results it can be concluded that this system can work according to its function and purpose.]]>
