Currently, in the era of 4.0, the development of computational-based statistical analysis techniques has been adapted in various fields, especially in the field of government. One of them is the use of the Support Smooth Vector Machine (SSVM) classification method for the classification of the Human Development Index in various regions in Indonesia.
The Human Development Index (HDI) is a parameter used to measure the achievement of human development based on a number of basic components of quality of life. As a measure of quality of life, HDI is built through a three-dimensional approach. These dimensions include long life and health; knowledge, and a decent life in a certain country or region. However, in classifying large amounts of data, errors often occur or produce low accuracy values, so that the results of the analysis cannot be used as a reference for judging decisions by the government.
First, I think the Support Smooth Vector Machine (SSVM) method is a solution for large data classification and can replace the previous method, namely the Support Vector Machine (SVM) by using the sigmoid integral neural network function. In my opinion, this method is suitable for classifying the human development index in Indonesia, considering that Indonesia is a large archipelago and the population in 2020 will reach 270 million people.
Other than that, The SSVM method also uses the karnel linear approach, namely a boundary plane matrix shows normal results, but cannot be differentiated twice so it cannot be used in the Newton method. According to the results of research by Pristiyani Darsyah (2019) regarding the classification of big data using the Newton-Armijo Algorithm method on SSVM. The results of these studies classification of the Human Development Index (HDI) used Kernel linear method shows the correct results for 3422 cities in Indonesia with a prediction accuracy of 84.77%. Whereas SSVM uses the Kernel Polynomial method of 61.65% and SSVM with the Radial Basis Function Kernel (RBF) method gives perfect prediction results of 100%.
In conclusion, I agree with the results of this study, that the Support Smooth Vector Machine (SSVM) method can be used for the classification of the human development index in Indonesia. In addition, in my opinion, SSVM specifically with the Radial Basis Function Kernel (RBF) method can be used as a reference for other classifications with large data sizes.
