Deep learning, specifically the Convolutional Neural Network (CNN), has been a significant technology tool for image processing and human health. CNNs, which mimic the working principles of the human brain, can learn robust representations of images. However, CNNs are susceptible to noise interference, which can impact classification performance. Choosing the right activation function can improve CNNs performance and accuracy. This research aims to test the accuracy of CNN with ResNet50, VGG16, and GoogleNet architectures combined with several activation functions such as ReLU, Leaky ReLU, Sigmoid, and Tanh in the classification of images that experience Poisson noise. Poisson noise is applied to each test data to evaluate CNN accuracy. The data used in this study consists of three scenarios of different numbers of classes, namely 3 classes, 5 classes, and 10 classes. The results showed that combining ResNet50 with the ReLU activation function produced the best performance in class recognition in each scenario of the number of classes experiencing Poisson noise interference. The model achieved 97% accuracy for 3-class data, 95% for 5-class data, and 90% for 10-class data. These results show that using ResNet50 with the ReLU activation function can provide excellent resistance to Poisson noise in image processing. It was found that as the number of classes increases, the accuracy of image recognition tends to decrease. This shows that the more complex the image classification task is with a larger number of classes, the more difficult it is for CNNs to distinguish between different classes. Doi: 10.28991/ESJ-2024-08-02-014 Full Text: PDF

One of the best-known and frequently used areas of Deep Learning in image processing is the Convolutional Neural Network (CNN), which has architectural designs such as Inceptionv3, DenseNet201, Resnet50, and MobileNet used in image classification and pattern recognition. Furthermore, the CNN extracts feature from the image according to the designed architecture and performs classification through the fully connected layer, which executes the Machine Learning (ML) algorithm tasks. Examples of ML that are commonly used include Naive Bayes (NB), k-Nearest Neighbor (k-NN), Support Vector Machine (SVM), and Decision Tree (DT). This research was conducted based on an AI model development background and the need for a system to diagnose COVID-19 quickly and accurately. The aim was to classify the aforementioned CNN models with ML algorithms and compare the models’ accuracy before and after Bayesian optimization using CXR lung images with a total of 2000 data. Consequently, the CNN extracted 80% of the training data and 20% for testing, which was assigned to four different ML models for classification with the use of Bayesian optimization to ensure the best accuracy. It was observed that the best model classification was generated by the MobileNetV2-SVM structure with an accuracy of 93%. Therefore, the accuracy obtained using the SVM algorithm is higher than the other three ML algorithms. Doi: 10.28991/HIJ-2023-04-03-05 Full Text: PDF

The problem of dimension reduction in multivariate data is how to obtain a smaller number of variables but still be able to maintain most of the information contained in the data. One method that can be used is principal component analysis (PCA). Principal component analysis (PCA) is a technique used to reduce the dimensions of data consisting of several dependent variables while maintaining the variance in the data. PCA can be used to stabilize measurements in statistical analysis, one of which is cluster analysis. Fuzzy clustering is a method of grouping based on membership values that include fuzzy sets as a basis for weighting for grouping. One method of fuzzy clustering is Fuzzy Subtractive Clustering (FSC). The method used in this study is PCA and FSC. The purpose of this study is to compare the most optimal cluster results using PCAFSC and FSC methods. The results obtained indicate that the clustering using the PCAFSC method is better than the FSC method.

Regression analysis is a method used to determine the relationship between one dependent variable and one or more independent variables. However, the existence of outliers in the 2018 Community Literacy Development Index data led to the application of statistical methods not sensitive to pencils for analysis. This was the reason for adopting robust regression methods, including the M, S, and MM estimations. Therefore, this research aims to compare these three estimates and select the one with the best estimate based on the parameter estimation model associated with the RSE and R2 values. Descriptive and inferential analysis with robust regression was used due to several outlier data and to provide good regression model results with unbiased values. It was discovered that the S-estimator and MM-estimator are the best methods because they have the most minor Residual Standard Error (RSE) of 1.856 and R2 of 0.9778.