IJCCS (Indonesian Journal of Computing and Cybernetics Systems)
Not a member yet
480 research outputs found
Sort by
Siamese-Network Based Signature Verification using Self Supervised Learning
Full text linkThe use of signatures is often encountered in various public documents ranging from academic documents to business documents that are a sign that the existence of signatures is crucial in various administrative processes. The frequent use of signatures does not mean a procedure without loopholes, but we must remain vigilant against signature falsification carried out with various motives behind it. Therefore, in this study, a signature verification system was developed that could prevent the falsification of signatures in public documents by using digital imagery of existing signatures. This study used neural networks with siamese network-based architectures that also empower self-supervised learning techniques to improve accuracy in the realm of limited data. The final evaluation of the machine learning method used gets a maximum accuracy of 83% and this result is better than the machine learning model that does not involve self-supervised learning methods
Audio-Visual CNN using Transfer Learning for TV Commercial Break Detection
Full text linkThe TV commercial detection problem is a hard challenge due to the variety of programs and TV channels. The usage of deep learning methods to solve this problem has shown good results. However, it takes a long time with many training epochs to get high accuracy. This research uses transfer learning techniques to reduce training time and limits the number of training epochs to 20. From video data, the audio feature is extracted with Mel-spectrogram representation, and the visual features are picked from a video frame. The datasets were gathered by recording programs from various TV channels in Indonesia. Pre-trained CNN models such as MobileNetV2, InceptionV3, and DenseNet169 are re-trained and are used to detect commercials at the shot level. We do post-processing to cluster the shots into segments of commercials and non-commercials. The best result is shown by Audio-Visual CNN using transfer learning with an accuracy of 93.26% with only 20 training epochs. It is faster and better than the CNN model without using transfer learning with an accuracy of 88.17% and 77 training epochs. The result by adding post-processing increases the accuracy of Audio-Visual CNN using transfer learning to 96.42%
Application of Extreme Learning Machine Method With Particle Swarm Optimization to Classify of Heart Disease
Full text linkPenyakit jantung koroner adalah tersumbatnya suplai darah jantung. Penyakit jantung adalah penyebab utama kematian di seluruh dunia. Berbagai faktor risiko berkontribusi terhadap penyakit jantung, termasuk merokok, gaya hidup tidak sehat, kolesterol tinggi, dan hipertensi. Dengan demikian, prediksi penyakit dapat dilakukan untuk mengidentifikasi individu yang berisiko guna mencegah peningkatan kematian akibat penyakit jantung. Penambangan data, khususnya metode Extreme Machine Learning (ELM), biasanya digunakan untuk tujuan ini. ELM adalah metode jaringan saraf dalam kecepatan pelatihan dan tidak memerlukan propagasi balik, dan menentukan jumlah node tersembunyi yang optimal dan mencapai hasil yang akurat tetap menjadi tantangan. Pada penelitian ini, ELM dengan Particle Swarm Optimization (PSO) diusulkan untuk mengoptimalkan klasifikasi penyakit jantung, yang bertujuan untuk mencapai hasil optimal dengan pembelajaran cepat. Penelitian ini mengikuti proses yang sistematis, termasuk pengumpulan data, preprocessing, pemodelan, dan evaluasi menggunakan analisis matriks konfusi. Hasil dan pembahasan menyajikan efektivitas metode yang diusulkan dengan mengevaluasi akurasi klasifikasi berdasarkan berbagai parameter, seperti ukuran populasi, jumlah node tersembunyi, dan iterasi. Temuan menunjukkan bahwa ELM dengan optimasi PSO dapat memberikan hasil klasifikasi yang akurat untuk diagnosis penyakit jantung, dengan tingkat akurasi yang menjanjikan
Convolutional Long Short-Term Memory (C-LSTM) For Multi Product Prediction
Full text linkThe retail company PT Terang Abadi Raya has a solid commitment to supporting distributors of LED lights and electrical equipment who have joined them, helping to spread their products widely in various regions. To face increasingly intense market competition, it is essential to produce high-quality products to win the competition and meet consumer demands. To achieve this, efficient production planning is necessary. The Convolutional Long Short-Term Memory (C-LSTM) method is used in this study to forecast product sales at PT Terang Abadi Raya. The research results show that C-LSTM has the potential to predict sales effectively. Evaluation is conducted using Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). The calculations reveal that the smallest values are obtained at epoch 10, with an MAE of 0.1051 and a MAPE of 22% in the testing data. For the cable data, the smallest values are found at epoch 100, with an MAE of 0.0602 and a MAPE of 44% in the testing data. The Long Short-Term Memory (LSTM) method with ten neurons produces the most minor errors during training
Improved LSTM Method of Predicting Cryptocurrency Price Using Short-Term Data
Full text linkAs cryptocurrencies develop, it cannot be denied that crypto prices are volatile. One of the influencing factors is the increasing volume of transactions which attracts the interest of researchers to conduct research in developing coin price predictions from cryptocurrencies. The method, algorithm and amount of data affect the prediction results. In this study, prediction modelling will be carried out using the LSTM method and short-term data. This study will conduct two experiments using the simple LSTM method and utilising multivariate time series with LSTM. The smallest predicted value is obtained using an 80/20 data allocation distribution scenario, input layer LSTM = 360, Epoch = 500, a Solana coin with RMSE = 0.111, R2 = 0.9962. It can be interpreted that short-term data can be used in making predictive models. Still, special attention needs to be paid to the characteristics of the dataset used and the modelling methodology, and it is hoped that the results of this study can be used in further research
Progressive Content Generation Based on Cyclic Graph for Generate Dungeon
Full text linkDungeon is level in game consisting collection of rooms and doors with obstacles inside. To make good level, takes a lot of time. With Procedural Content Generation (PCG), dungeons can be created automatically. One of the approaches in PCG to create levels is progressive. Progressive approach produces timeline as representation of the interactions in the game. Timeline representation that is in the form of one straight line is good for endless runner, but for dungeon, the levels are linear. In this research, the timeline is changed to cyclic graph. Cyclic graph is formed using graph grammar algorithm. This research aims to build dungeon that has not linear and minimal dead ends. To eliminate linearity in dungeons, branching in dungeons needs to be formed. The steps carried out in this research are designing graph grammar rules, generating population of graphs, evaluating graphs with fitness values, and building dungeons. Four functions are used to determine the fitness value: shortest vertices, average duration, replayability, and variation. Dungeons produced with progressive approach manage to minimize linearity in dungeons. Dungeon formation is very dependent on the rule grammar that forms it. With the evaluation process, linear dungeons resulting from grammar rules can be minimized
Exploring Pre-Trained Model and Language Model for Translating Image to Bahasa
Full text linkIn the last decade, there have been significant developments in Image Caption Generation research to translate images into English descriptions. This task has also been conducted to produce texts in non-English, including Bahasa. However, the references in this study are still limited, so exploration opportunities are open widely. This paper presents comparative research by examining several state-of-the-art Deep Learning algorithms to extract images and generate their descriptions in Bahasa. We extracted images using three pre-trained models, namely InceptionV3, Xception, and EfficientNetV2S. In the language model, we examined four architectures: LSTM, GRU, Bidirectional LSTM, and Bidirectional GRU. The database used was Flickr8k which was translated into Bahasa. Model evaluation was conducted using BLEU and Meteor. The performance results based on the pre-trained model showed that EfficientNetV3S significantly gave the highest score among other models. On the other hand, in the language model, there was only a slight difference in model performance. However, in general, the Bidirectional GRU scored higher. We also found that step size in training affected overfitting. Larger step sizes tended to provide better generalizations. The best model was generated using EfficientNetV3S and Bidirectional GRU with step size=4096, which resulted in an average score of BLEU-1=0,5828 and Meteor=0,4520
Information Retrieval for Early Detection of Disease Using Semantic Similarity
Full text link The growth of medical records continues to increase and needs to be used to improve doctors' performance in diagnosing a disease. A retrieval method returns proposed information to provide diagnostic recommendations based on symptoms from medical record datasets by applying the TF-IDF and cosine similarity methods. The challenge in this study was that the symptoms in the medical record dataset were dirty data obtained from patients who were not familiar with biological terms. Therefore, the symptoms were matched in the medical record data with the symptom terms used in the system and from the results, data augmentation was carried out to increase the amount of data up to about 3 times more. In the TF-IDF the highest accuracy with is only , while after augmentation of the test data, the accuracy becomes . The highest accuracy results with the same value using the cosine similarity method is and with the augmented test data accuracy increasing to . From this study it was concluded that a system with sufficient and relevant input of symptoms would provide a more accurate disease prediction. Prediction results using the TF-IDF method with are more accurate than predictions using the cosine similarity method
Autism Spectrum Disorder (ASD) Identification Using Feature-Based Machine Learning Classification Model
Full text linkAutism Spectrum Disorder (ASD) is a developmental disorder that impairs the development of behaviors, communication, and learning abilities. Early detection of ASD helps patients to get beter training to communicate and interact with others. In this study, we identified ASD and non-ASD individuals using machine learning (ML) approaches. We used Gaussian naive Bayes (NB), k-nearest neighbors (KNN), random forest (RF), logistic regression (LR), Gaussian naive Bayes (NB), support vector machine (SVM) with linear basis function and decision tree (DT). We preprocessed the data using the imputation methods, namely linear regression, Mice forest, and Missforest. We selected the important features using the Simultaneous perturbation feature selection and ranking (SpFSR) technique from all 21 ASD features of three datasets combined (N=1,100 individuals) from University California Irvine (UCI) repository. We evaluated the performance of the method's discrimination, calibration, and clinical utility using a stratified 10-fold cross-validation method. We achieved the highest accuracy possible by using SVM with selected the most important 10 features. We observed the integration of imputation using linear regression, SpFSR and SVM as the most effective models, with an accuracy rate of 100% outperformed the previous studies in ASD predicito
Simulation Technique in Determining Student Attendance Using The Monte Carlo Method
Full text linkIn lectures, attendance is one of the assessment points that play an important role in determining a student's graduation. When a student is in the upper semester their attendance rate at lectures starts to decrease. The attendance prediction simulation is an estimate of the calculation of student attendance in lectures. This type of research is quantitative research using data collection techniques using observation and documentation study. In the process of analysis, the observed data were attendance data of 5th-semester computer engineering study program students and a sample of 83 people as research subjects. The stages of the monte carlo simulation are used: Determining variable frequency; Calculating cumulative probabilities; Determining random number intervals; Creating a simulation to determine student attendance; Generating random numbers; Make a simulation of the experimental circuit. The simulation is carried out by comparing and entering random numbers that have been generated into a comparison simulation of attendance and absence data for 5th-semester computer engineering study program students at the STMIK Agamua Wamena Papua Campus, starting from October 3 to October 31, 2022. Based on a series of experimental data that has The simulation results obtained predicted attendance and absence of computer engineering study program students at the STMIK Agamua Wamena campus from November 7 to December 19, 2022 with an average attendance of above 50%