Various machine learning algorithms are applicable in the diagnosis and prediction of diabetes mellitus. Since each of the classifiers has a different implementation matrix, their prediction accuracies vary. Therefore, this study focused on comparing the accuracy of various machine learning algorithms in predicting diabetes mellitus. The algorithms include ID3, C4.5, CART, CHAID, Gradient Boosting, Random Forest, and AdaBoost. The implementation of the algorithms was done using the Chefboost decision tree-based framework in Python. The obtained results they concluded that the accuracy of prediction for all the implemented algorithms is satisfactorily high. However, tree-based classification algorithm ID3 had a higher prediction accuracy than all the others.

Advanced Technology like Internet of Things (IoT) enables users to control hardware devices through a software over a wide coverage area at very fast speed with little source of energy. This energy places an important role in household appliances, industries, agricultural machinery and so on. Controlling the energy efficiently for appliances in the home is very important too. In this work, home appliances are controlled through the internet. The design developed into a system that can control several electrical loads which were connected at different terminals. Upon connection then User interface were developed to interact with the system that allows user to easily control these home appliances through the internet. The system when routed through the IP address enable easy access to the appliances at home from distant locations. Thus, the time delay inherent in-Home automation systems technologies that includes Bluetooth, Zig bee and Z-wave uses remote control either through sending (SMS) are eliminated in IoT devices and designs based on the technology in question. However, the user can control the appliances with great ease thereby conserving the consumption of energy control proportional as a result.