<![CDATA[Infant nutritional status is very important to be known by the parents, because there are still many malnutrition cases of children under five in Indonesia that is unsolved. It considered because malnutrition not only by physical condition that affect. Based on these problems, a system of infant nutrition status have been made using K-Nearest Neighbor method based on embedded system using gender, age and body weight parameters of infant. The value of K that would be applied to the method needs to be tested to get the best value of K for the system. For readings body weight using a HX711 circuit module, a load cell sensor, connected to NodeMCU ESP8266 in order to send the data wirelessly to a computer / PC as a classifier parameter. The input are gender and age as requiredment for classification, if all three parameters are met the nutritional status can be displayed on web and saved in database as the archieve. From the analysis that has conducted, it can be concluded that the functional testing on the weight sensor has 97,23% accuracy, while the functional testing of basis data storage has 100% conformity. K value has the highest percentage of accuracy when K = 5 and k = 6 with 62.50%. While overall testing of the baby's nutritional status classification system yielded 97,14% accuracy.]]>
Copyrights © 2017
