Handwriting recognition in Javanese script is not widely developed with deep learning (DL). Previous DL and machine learning (ML) research is generally limited to basic characters (Carakan) only. This study proposes a deep learning model using a custom-built convolutional neural network to improve recognition accuracy performance and reduce computational costs. The main features of handwritten objects are textures, edges, lines, and shapes, so convolution layers are not designed in large numbers. This research maximizes optimization of other layers such as pooling, activation function, fully connected layer, optimizer, and parameter settings such as dropout and learning rate. There are eleven main layers used in the proposed custom convolutional neural network (CNN) model, namely four convolution layers+activation function, four pooling layers, two fully connected layers, and a softmax classifier. Based on the test results on the Javanese script handwritten image dataset with 120 classes consisting of 20 basic character classes and 100 compound character classes, the resulting accuracy is 97.29%.
10 May 2021 in Journal of Applied Intelligent System
This study focuses on the analysis of sentiments on Indonesian twitter data. Twitter data on Indonesian simultaneous pilkada used to get its sentiments using Naïve Bayes Classifier method as a method of classification and K-means method to get Label on the data train process. Combining the two methods is expected to get high accuracy results. The results obtained from the research shows a pretty good accuracy of 74.5%.
23 April 2024 in Scientific Journal of Informatics
Purpose: Hiragana image classification poses a significant challenge within the realms of image processing and machine learning. Despite advances, achieving high accuracy in Hiragana character recognition remains elusive. In response, this research attempts to enhance recognition precision through the utilization of a Convolutional Neural Network (CNN). Specifically, the study explores the efficacy of three distinct optimizers like Adam, Stochastic Gradient Descent with Momentum (SGDM), and RMSProp in improving Hiragana character recognition accuracy. Methods: This research adopts a systematic approach to evaluate the performance of a Convolutional Neural Network (CNN) in the context of Hiragana character recognition. A meticulously prepared dataset is utilized for in-depth testing, ensuring robustness and reliability in the analysis. The study focuses on assessing the effectiveness of three prominent optimization methods: Stochastic Gradient Descent (SGD), RMSProp, and Adam. Result: The results of the model performance evaluation show that the highest accuracy was obtained from the RMSP optimizer with an F1-Score reaching 99.70%, while the highest overall accuracy was 99.87% with the Adam optimizer. The analysis is carried out by considering important metrics such as precision, recall, and F1-Score for each optimizer. Novelty: The performance results of the developed model are compared with previous studies, confirming the effectiveness of the proposed approach. Overall, this research makes an important contribution to Hiragana character recognition, by emphasizing the importance of choosing the right optimizer in improving the performance of image classification models.
5 August 2024 in Scientific Journal of Informatics
Purpose: This research examines how well Schur decomposition-based steganography can hide data in digital images without being noticed, while also keeping the image quality high and keeping the hidden information safe. Methods: The study starts by choosing regular test images (Lena, Plane, Peppers, Cameraman, Baboon) to use for hiding messages in. The Schur decomposition is used to hide information within images in a subtle way. To test how well the new method works, we added Gaussian noise and Salt & Pepper noise after embedding. The quality of the image is determined by looking at the Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index (SSIM) metrics. Result: The research shows that Schur decomposition results in very good SSIM values (greater than 0.92) and high PSNR scores (as high as 90.27 dB) for various image sizes (64x64, 128x128, 256x256). This means that the quality of the images is not greatly reduced even after steganography is applied. Novelty: This research introduces a unique use of Schur decomposition for hiding data in images without affecting their quality. The study highlights how this method can securely hide information in digital media, which could be really useful for improving steganography techniques in the future. Future studies should concentrate on making improvements to Schur decomposition-based steganography, especially for bigger images. One possibility is to create adaptive methods that can change how images are hidden based on their content. This could make it harder for others to detect and analyze hidden information in the images.
Overlap is one of the problems in the segmentation process. This research proposes a mini-imposition method to optimize the watershed segmentation and extended-maxima transformation methods. Morphological operations using the bwareopen function are also added at the end of the watershed segmentation process. This method is implemented to create a counter object on a migratory bird image that has many overlapping objects. The test results of this method are proven to improve the performance of the previous method and get an average value that is very satisfying with an average value of accuracy of 0.94, recall of 0.95, and precision of 0.99.
12 June 2023 in Spekta (jurnal Pengabdian Kepada Masyarakat : Teknologi Dan Aplikasi)
Background: Community-based tourism is increasingly being considered a more sustainable alternative. Tourism Villages can be a step towards realizing sustainable tourism. This is because the tourism village has the opportunity to open up as much participation as possible for the local community. However, this does not mean anything if managers, including accommodation providers, need qualified organizational and financial management skills. Contribution: This activity aims to provide knowledge to homestay managers in Candirejo Village to properly manage business finances based on SAK EMKM (Standar Akuntansi Keuangan Entitas Mikro, Kecil, dan Menengah) Method: The method used by researchers is through training and empowerment of homestay managers Results: The result of this activity is an increased understanding of SAK EMKM in Candirejo Village. After the training, homestay managers have also started implementing SAK EMKM in their financial reports. As a result, the financial position can be monitored, financial leaks can be minimized, and homestay business management can be carried out more effectively and efficiently. Conclusion: The activities that were held went well. This is known from the excellent understanding of SAK EMKM after the training was given