IJCCS (Indonesian Journal of Computing and Cybernetics Systems)
Not a member yet
480 research outputs found
Sort by
Face Image Generation and Enhancement Using Conditional Generative Adversarial Network
Get PDFThe accuracy and speed of a single image super-resolution using a convolutional neural network is often a problem in improving finer texture details when using large enhancement factors. Some recent studies have focused on minimal mean square error, resulting in a high peak signal to noise ratio. Generally, although the peak signal to noise ratio has a high value, the output image is less detailed. This shows that the determination of super-resolution is not optimal. Conditional Generative Adversarial Network based on Boundary Equilibrium Generative Adversarial Network, by combining Mean Square Error Loss and GAN Loss as a loss function to optimize the super-resolution model and produce super-resolution images. Also, the generator network is designed with skip connection architecture to increase convergence speed and strengthen feature distribution. Image quality value parameters used in this study are Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index (SSIM). The results showed the highest image quality values using dataset validation were 26.55 for PSNR values and 0.93 for SSIM values. The highest image quality values using the testing dataset are 24.56 for the PSNR value and 0.91 for the SSIM value
Lampung Script Recognition Using Convolutional Neural Network
Get PDFThe Lampung script is often used in writing words in Lampung language. The Lampung language itself is used by native Lampung people and people who learn Lampung language. The Lampung script is difficult to learn because there are many combinations of parent characters and subletters. CNN is a method in the field of object recognition that has a specific layer, namely a convolution layer and a pooling layer that allows the feature learning process well. Handwriting recognition as in character recognition in MNIST, CNN produces better performance compared to other methods. From the advantages of CNN, the CNN method with DenseNet architecture was chosen as the best architecture to recognize each Lampung script. In this study, there are 2 main processes, namely preprocessing, and recognition. This study succeeded in applying the CNN method which can recognize Lampung script. The dataset is divided into 4 groups of characters that have different sounds. First, the parent character data get 98% accuracy. Second, the parent letter data with the above letters get 98% accuracy. Third, the parent character data with the sub-letters on the side get 98% accuracy. Fourth, the parent letter data with the lower letters get 97% accuracy
Sentiment Analysis Using Backpropagation Method to Recognize the Public Opinion
Get PDF Improve the service quality of tourism actors by conducting sentiment analysis on digital platforms owned by tourism businesses and collecting negative sentiments to improve the quality of services from companies owned by tourism businesses. The growth of the hospitality industry in Indonesia is experiencing rapid growth every year. The tourism industry, part of the hospitality industry, also does not escape the influence of positive and negative sentiments. One method to perform accurate sentiment analysis is Backpropagation Neural Network. Based on the results of tests on the neural network, the best accuracy is obtained when using one hidden layer with the first layer of 10 neurons. The learning rate is 0.000002, where the accuracy is 71.630%. More epochs do not guarantee better accuracy. Based on the results of the research that has been done, suggestions for further researchers are to analyze the review dataset processing method so that it gets a cleaner dataset and is expected to improve better accuracy
Modification Weight Criteria With Webbed Model For Selection Artist Music Festival Using Analytical Hierarchy Process (AHP)
Get PDFThe process of selecting from many alternatives to the criteria is a decision that is often determined in decision making. The criteria for which criteria can consist of many attributes are used by decision-makers in making the selection or are called multi-criteria decision making (MADM ). Determining the Artist Music Festival at an event has quite a complicated difficulty, because the assessment of the criteria is heterogeneous. The spider web's approach to integrating criteria results in the selection of the artist form that attracts the most attention, and public interest. Model AHP use of multi-attributes is used in selecting artists to perform at music festivals, selecting artists using criteria, namely Number of Followers (C1), stamp C2, Average Popular Tracks (C3), Average Youtube Viewers (C4), and the price of the artist (C5). Data on the number of followers, popularity, average popular tracks, and average YouTube viewers were obtained using the Spotify and Youtube APIs. The settlement method applied is the Analytical Hierarchy Process (AHP) and the Rating Scale algorithm, with an alternative using five samples of Indonesian artists. The research results are expected to provide recommendations for artists as performers in the music festival
Financial Distress Prediction with Stacking Ensemble Learning
Get PDFPrevious studies have used financial ratios extensively to build their predictive model of financial distress. The Altman ratio is the most often used to predict, especially in academic studies. However, the Altman ratio is highly dependent on the validity of the data in financial statements, so other variables are needed to assess the possibility of manipulation of financial statements. None of the previous studies combined the five Altman Ratios with the Beneish M-Score. We use Stacking Ensemble Learning to classify crisis companies and perform a comprehensive analysis. This insight helps the investment public make lending decisions by mixing all the financial indicator information and assessing it carefully based on long-term and short-term conditions and possible manipulation of financial statements
On the Design of a Blockchain-based Fraud-prevention Performance Appraisal System
Get PDF The job recruitment process takes a lot of process and number of documents. It is very well known for applicants to exaggerated and falsify their work history data. It may put a company at legal risk and significant commercial losses. Generally, company use third-party to verify applicant’s work history data which is time-consuming and costly. It also makes companies relies on third-party which may not trustworthy and cause several other risks. Generally, experience letters is used as a proof of work history documents of employee. However, the process of publishing an experience letter may contain conflict of interest between company and employee. Yet, publishing an experience letter is not mandatory in several places. In this research, we propose a system to verify applicant’s work history data by using performance appraisal as proof of work history and utilizing Blockchain to provide secure system, tampered-proof and real-time verification. The proposed approach also minimizes trust issues and privacy of data sharing by adding encryption and digital signature schema using Elliptic Curve Cryptography (ECC) algorithm. Furthermore, we have implemented a prototype to demonstrate how the proposed system work using a Quorum-based consortium blockchain
Tegal Tourism Object Selection Decision Support System Using Fuzzy Logic
Get PDFThere are many agencies that have databases but are left without proper management. For example, in tourist attractions in Kota & Kab. Tegal, along with the rapid development of tourism technology, the tourism industry requires the tourism industry to apply information technology to provide convenience for tourists to find out tourist areas according to the cost, and the distance of the tourist attractions entered. The provision of tourism information helps tourists to consider and make decisions to travel. Tahani Fuzzy Logic was chosen because the concept of Fuzzy logic is easy to understand, flexible, and because the Tahani logic method is a form of decision support where the main tool is functional with the main input criteria determined by the user/tourist. This system was implemented using web programming and MySQL database, where the variables to be considered are Type of Tour, Number of Facilities, Price of Tour Tickets, Number of Tourist Visitors, Travel Distance from City Center. The results of this study are a decision support system for tourism selection in Tegal using Fuzzy Tahani which can recommend tourist attractions in Tegal which are determined by tourists depending on tourist criteria based on the firestrength of the selected variable
Identification of Incung Characters (Kerinci) to Latin Characters Using Convolutional Neural Network
Get PDFIncung script is a legacy of the Kerinci tribe located in Kerinci Regency, Jambi Province. On October 17, 2014, the Incung script was designated by the Ministry of Education and Culture as an intangible heritage property owned by Jambi Province. But in reality, the Incung script is almost extinct in society. This study aims to identify the characters of the Incung (Kerinci) script with the output in the form of Latin characters from the Incung script. The classification method used is the Convolutional Neural Network (CNN) method. The dataset used as many as 1400 incung character images divided into 28 classes. In this study, an experiment was conducted to obtain the most optimal model. Showing the results using the CNN method during the training process that the accuracy of the training data reaches 99% and the accuracy of the testing data reaches 91% by using the optimal hyperparameters from the tests that have been done, namely batch size 32, epoch 100, and Adam's optimizer. It evaluates the CNN model using 80 images in words (a combination of several characters) with 4 test scenarios. It shows that the model can recognize image data from scanning printed books, digital writing test data, test data with images containing more than two characters, and check images with different font size
Increasing Performance of Multiclass Ensemble Gradient Boost uses Newton-Raphson Parameter in Physical Activity Classifying
Get PDFThe sophistication of smartphones with various sensors they have can be used to recognize human physical activity by placing the smartphone on the human body. Classification of human activities, the best performance is obtained when using machine learning methods, while statistical methods such as logistic regression give poor results. However, the weakness of the logistic regression method in classifying human activities is corrected by using the ensemble technique. This paper proposes to apply the Multiclass Ensemble Gradient Boost technique to improve the performance of the Logistic Regression classification in classifying human activities such as walking, running, climbing stairs, and descending stairs. The results show that the Multiclass Ensemble Gradient Boost Classifier by Estimating the Newton-Raphson Parameter succeeded in improving the performance of logistic regression in terms of accuracy by 29.11%
Topic Modeling on Online News.Portal Using Latent Dirichlet Allocation (LDA)
Get PDFThe amount of News displayed on online news portals. Often does not indicate the topic being discussed, but the News can be read and analyzed. You can find the main issues and trends in the News being discussed. It would be best if you had a quick and efficient way to find trending topics in the News. One of the methods that can be used to solve this problem is topic modeling. Theme modeling is necessary to allow users to easily and quickly understand modern themes' development. One of the algorithms in topic modeling is the Latent Dirichlet Allocation (LDA). This research stage begins with data collection, preprocessing, n-gram formation, dictionary representation, weighting, topic model validation, topic model formation, and topic modeling results. Based on the results of the topic evaluation, the. The best value of topic modeling using coherence was related to the number of passes. The number of topics produced 20 keys, five cases with a 0.53 coherence value. It can be said to be relatively stable based on the standard coherence value