Analysis of Membership Function in the Implementation of Neuro-Fuzzy System for Prediction of Depressive Lexicons
Pilita A. Amahan, Mia V. Villarica, Albert Alcause Vinluan · 2021
The soft computing environment does not stop making predictions to save lives. Its calculations, though not precise can still provide solutions to the complex scenarios of computational problems. Over the years, robustness and the accuracy of results became the remarkable challenge despite of the huge number of studies relating to it. In this case, the Neuro fuzzy system is one amongst the model proven to be used for prediction and has been applied to various fields of education, business, engineering and even in health. The Neuro fuzzy has its capability to adopt various strategies like the mix and match approach with different parameters involved during the experimentation. However, there is still a lack of support as to how this membership function could be able to produce a precise result that leads to different views towards soft computing. In this work, we gave emphasize on how the membership function works with NFS, which is essential in preparing data towards the prediction of depressive lexicons. Classifiers like Naïve Bayes, Simple Logistics, and Neuro fuzzy through the Multilayer Perceptron (MLP) of Weka tool have been used for the experimentation of the study. This study recommends to use the process of the analysis of membership function and carefully validate parameters from the first layer towards the third layer of the NFS since the strength of the membership function actively happens here.