Articles

Access the latest knowledge in applied science, electrical engineering, computer science and information technology, education, and health.

Advanced Search
Filter Icon

Filters article

Years

FAQ Arrow
Single Year
Range Year
0
0

Source Title

FAQ Arrow

Authors

FAQ Arrow

31,377 Article Results

Image enhancement combined with EfficientNet-B7 for grading classification of diabetic retinopathy

10.11591/ijai.v15.i3.pp2554-2566
Nina Sevani , Edy Kristianto , Albert Salomo
Diabetic retinopathy (DR) is a complication caused by poorly managed diabetes that affects the eyes. According to the World Health Organization (WHO), 422 million people worldwide have suffered from DR in the past ten years. Manual detection using retinal fundus images is time-consuming and requires experienced ophthalmologists. This study proposes a deep learning method using the pre-trained model EfficientNet-B7 to identify this disease automatically. Five levels of DR will be classified: no-DR, mild-DR, moderate-DR, severe-DR, and proliferative-DR. The model was trained using "APTOS 2019 blindness detection" dataset, and image augmentation was performed. Image segmentation techniques such as contrast limited adaptive histogram equalization (CLAHE) and real enhanced super resolution generative adversarial network (Real-ESRGAN) were applied during preprocessing to improve the model's accuracy significantly. The implementation of CLAHE resulted in the validation accuracy improvement from 76.6% to 83.4% compared to no segmentation, while the combination of Real-ESRGAN and CLAHE increased the accuracy to 93.7%. Future research can explore the combination of CLAHE with other image processing techniques apart from the Real-ESRGAN model.
Volume: 15
Issue: 3
Page: 2554-2566
Publish at: 2026-06-01

Can machines imagine? Critical thinking and cultural reasoning in multimodal-multilingual AI

10.11591/ijict.v15i2.pp823-838
Mohammad Awad AlAfnan , Siti Fatimah MohdZuki , Shefa Mohammad AlAfnan
Effective communication across languages and cultures is essential in today’s interconnected world. Multimodal-multilingual language models (MMMLMs) aim to advance this goal by integrating text, speech, and visual understanding across diverse linguistic contexts. This study evaluates four leading MMMLMs-GIT, mPLUG, CLIP, and Whisper + GPT-4V-on cross lingual and cross-modal tasks, including image captioning, visual question answering, speech-to-image generation, and idiomatic translation. Performance was assessed in high-resource (English, Arabic), medium resource (Malay), and low-resource (Macedonian) settings. Results show strong performance in structured tasks but notable limitations in cultural reasoning, figurative language interpretation, and semantic grounding in low-resource environments. GIT delivered the most consistent multilingual results, while Whisper + GPT-4V excelled in fluency yet lacked cultural sensitivity. To address these gaps, the study proposes culturally informed evaluation protocols that integrate quantitative metrics such as BLEU, CIDEr, and F1 with qualitative, community-centered approaches. These include cross-cultural annotation panels, inter-rater reliability validation using Cohen’s kappa, and a novel “cultural fidelity” metric to measure alignment with culturally specific norms. The findings emphasize the need for inclusive datasets, ethical development, and interdisciplinary collaboration to ensure MMMLMs support equitable and culturally aware global communication.
Volume: 15
Issue: 2
Page: 823-838
Publish at: 2026-06-01

Predicting battery life performance using artificial intelligence techniques in electric vehicles

10.11591/ijict.v15i2.pp805-812
Debani Prasad Mishra , Munavath Pavan Kalyan , Shivam Tyagi , Piyushjeet Piyushjeet , Shiv Grover , Surender Reddy Salkuti
Electric vehicles’ (EVs’ performance and sustainability are significantly influenced by the efficiency and lifespan of their lithium-ion batteries. This paper explores the critical factors affecting battery degradation, focusing on parameters such as charge cycles, thermal management, and voltage dynamics. Utilizing a dataset of 14 batteries, the study employs data-driven machine learning (ML) to predict the remaining useful life (RUL) of batteries. The ensemble-based regression model demonstrated superior predictive accuracy through comprehensive analysis, achieving R² values of 97.89% for training and 94.69% for testing. Feature importance analysis identified cycle index (CI) as the most critical determinant of battery health, followed by discharge time and voltage stability. Visualizations, including correlation heatmaps and residual plots, validate the robustness of the selected model. Additionally, sustainable charging strategies, such as steady current-steady voltage (also known as CC-CV), are highlighted for their role in enhancing battery longevity. This research offers actionable insights into battery management systems, providing a robust foundation for predictive maintenance and the development of sustainable electric mobility solutions.
Volume: 15
Issue: 2
Page: 805-812
Publish at: 2026-06-01

Machine learning approaches for anomaly detection of Jakarta air quality index

10.11591/ijai.v15.i3.pp2543-2553
Muhammad Rizky Nurhambali , Yenni Angraini , Anwar Fitrianto
Anomalies in time series data are observations that deviate markedly from surrounding values or overall patterns. Air quality index (AQI) data, which vary over time, provide a suitable context for anomaly detection. Time series anomaly detection can be done with machine learning approaches like long short-term memory (LSTM) and extreme gradient boosting (XGBoost). These methods have advantages over conventional methods in handling nonlinearity and large data dimensions. This study compares LSTM and XGBoost methods for detecting anomalies in Jakarta's hourly AQI data. The dataset was obtained from the AirNow website and covers the period from January 1, 2018, to December 31, 2023. Anomalies in the observed data were labeled using moving range (MR) (2) and (3) approaches with three and four-sigma thresholds, and feature engineering (FE) was applied to improve model performance. The results indicate that LSTM is more suitable than XGBoost for forecasting and classification tasks in AQI data. LSTM achieved an average mean absolute percentage error (MAPE) of 10.3840%, a root mean square error (RMSE) of 10.5913, and a balanced accuracy (BACC) of 0.9424 under MR (2) labeling with the four-sigma rule. The anomalies detected mostly occurred between 21:00 and 09:00 and during the rainy season.
Volume: 15
Issue: 3
Page: 2543-2553
Publish at: 2026-06-01

Brain tumor detection using VGG-16 model

10.11591/ijai.v15.i3.pp2337-2346
Aicha Oussous , Abderrahmane Ez-zahout , Soumia Ziti
Research in medical image analysis, specifically through deep convolutional networks, addresses the challenges of manually analyzing large magnetic resonance imaging (MRI) image volumes for brain tumor detection. The manual analysis is time-consuming, tedious, and prone to inaccuracies due to subtle visual similarities between normal tissue and tumor cells. This research aims to automate tumor detection, increasing accuracy and efficiency in medical treatments. This study aimed to develop a model capable of classifying brain tumors 2D MRI images, and the convolutional neural network (CNN)-based model successfully achieved an accuracy of 99.21% but suffered from noticeable Overfitting. Implementing the independent tests set and early stopping mitigated this issue, making the model more reliable for production deployment and demonstrating its potential in supporting physicians in detecting brain tumors, thereby enhancing treatment efficiency. The use of Python, TensorFlow, and Keras facilitated the development of the proposed solution, focusing on a diverse set of MRI images with varying tumor sizes, locations, shapes, and intensities.
Volume: 15
Issue: 3
Page: 2337-2346
Publish at: 2026-06-01

Evaluating user experience of a mobile website and redesigning its user interface using goal-directed design method

10.11591/ijict.v15i2.pp634-643
Aang Subiyakto , Muhammad R. Alghifari , Nuryasin N. , Muhammad Q. Huda , Nashrul Hakiem , Viva Arifin , Dwi Yuniarto , Hadi Rahman , Thosporn Sangsawang , Naeem Atanda Balogun
This study evaluated the usability of the user interface (UI) of a mobile website using its user experience (UX) perspectives. The website serves as an information portal intended for access via smartphones and other handheld devices. The objective of the study was to assess the usability of its current interface, redesign it using the goal-directed design (GDD) method, and compare the usability performance before and after the redesign. The study was conducted in five main steps using the cognitive walkthrough, think-aloud, post-study system usability questionnaire (PSSUQ), and interview techniques with five representative participants and 50 respondents. The most important findings of the study were that the redesigned mobile website showed improved usability of the website, as indicated by increased effectiveness and efficiency values, enhanced PSSUQ satisfaction scores, and more positive user feedback.
Volume: 15
Issue: 2
Page: 634-643
Publish at: 2026-06-01

Intelligent home automation framework using sensor fusion and machine learning for energy efficiency and thermal comfort

10.11591/ijict.v15i2.pp545-552
Franklin Ovuolelolo Okorodudu , Gracious Chukwuweike Omede , Etinosa Eugene Osawe
This paper presents an innovative, intelligent home automation framework integrating sensor fusion and machine learning to promote energy efficiency and thermal comfort in residential settings. Utilising low-cost hardware such as the Arduino Uno R3, passive Infrared (PIR) sensors, KY-018 photoresistors, and KY-028 temperature sensors, the system achieves a human presence detection accuracy of 95.3% via a random forest classifier. Over a three-month period, testing in several homes showed that the system is 99.7% reliable, responds in 1.2 seconds, and costs 85% less than commercial options. This research lays the groundwork for sustainable smart homes by providing a mathematical model for optimizing energy use and a unified modeling language (UML) model of the system architecture. These results show how important it is to have open-source technology that is cheap and could help smart building systems spread around the world. The study utilized a controlled experimental design featuring five families, with sensor data gathered at 10-second intervals over a three-month period. A random forest classifier trained on 10,000 labeled data points could correctly guess whether or not a person was present 94.8% of the time and 95.7% of the time. The framework is useful because it combines cheap sensors with a lightweight machine-learning pipeline that can work on small microcontrollers. This solves the long-standing problem of the cost performance gap seen in prior smart-home deployments.
Volume: 15
Issue: 2
Page: 545-552
Publish at: 2026-06-01

Enhancing Bitcoin price forecasting: a comparative analysis of advanced time series models with hyperparameter optimization

10.11591/ijict.v15i2.pp535-544
Amine Batsi , Mohamed Biniz , Rachid El Ayachi
This paper evaluates state-of-the-art time series forecasting to predict next day Bitcoin prices via distinct architectures and methodologies in a real-time setting. We study six advanced models, KAN, TimesNet, NBEATS, NHITS, PatchTST and BiTCN, applied to a Jan 1, 2023, to Dec 1, 2024. We simulate real world applications via a rolling forecast strategy, in which we predict daily prices from the most recent data. The dataset consists of daily Bitcoin closing prices and data preprocessing and integrity checks for its constituent data. Additionally, rigorous accuracy and reliability were investigated using performance metrics such as the MAE, RMSE, MAPE, and R². NBEATS and NHITS were the top performers, achieving an R² score of 0.967, explaining complex patterns in volatile cryptocurrency data. The specific importance of model architecture and further hyperparameter optimization in achieving higher forecasting accuracy is highlighted in this study. The practical implications of these findings for the advancement of time series forecasting in financial markets are leveraged here, where timely and accurate forecasts are critical.
Volume: 15
Issue: 2
Page: 535-544
Publish at: 2026-06-01

Deep learning-based optimization techniques for network lifetime enhancement in wireless sensor networks

10.11591/ijict.v15i2.pp623-633
Abhay Raghunath Gaidhani , Amol D. Potgantwar
Wireless sensor networks (WSNs) are integral to applications like environmental monitoring, healthcare, and surveillance, yet they face the critical challenge of limited energy resources, which shortens the network's operational lifespan. Addressing this issue, this paper explores deep learning-based optimization techniques as a solution to enhance network lifetime by efficiently managing energy consumption. We present a detailed review of the existing challenges in WSNs and examine various deep learning methods, including neural networks, deep reinforcement learning (DRL), and generative adversarial networks, specifically tailored for WSN optimization. In this study, we introduce a new reinforcement learning (RL) based optimization algorithm to prolong the network lifetime. The proposed technique is intended to smartly distribute the energy consumption among the network elements, leading to desirable performance with regard to the battery lifetime. The paper ends with a summary of design aspects and future research directions to improve the WSN performance further based on deep learning.
Volume: 15
Issue: 2
Page: 623-633
Publish at: 2026-06-01

Artificial intelligence-based risk assessment in agro-industry using supervised neural networks

10.11591/ijai.v15.i3.pp2260-2268
Imam Santoso , Izzum Wafi'uddin , Naila Maulidina Lu'ayya , Annisa'u Choirun , Siti Asmaul Mustaniroh , Dodyk Pranowo , Ainur Rofiq
The coffee supply chain involves high production volumes, complex multi actor interactions, and increasing sustainability requirements, yet remains highly vulnerable to risks dimension. This study aims to develop and evaluate a decision-support framework that improves the accuracy and consistency of sustainability risk classification in the coffee supply chain. The proposed framework integrates failure mode and effect analysis (FMEA) with a supervised artificial neural network (ANN) using backpropagation (BP) to enable data-driven and adaptive risk assessment. Empirical data was collected from 55 respondents, resulting in the identification of 35 supply chain risk factors. These data were used to train and validate an ANN-based classification model implemented in a Python environment, with standard preprocessing and stratified data partitioning to ensure robustness. The ANN classified risks into five categories using supervised learning. The results demonstrate strong predictive performance, achieving overall accuracy of 98.97%, with precision, recall, and F1-scores exceeding 96.8% across all risk classes. Confusion matrix analysis confirms reliable generalization and minimal misclassification. The findings indicate that integrating FMEA with ANN-BP significantly enhances risk classification compared to conventional qualitative approaches. The proposed framework provides a scalable and reliable decision-support tool for dynamic risk scoring, supporting enhancement of sustainable practices in agro-industrial coffee supply chains.
Volume: 15
Issue: 3
Page: 2260-2268
Publish at: 2026-06-01

Utilization of depth-wise and spatially separable convolutional network fusion for classification of white blood cells

10.11591/ijai.v15.i3.pp2595-2605
Firas Muneam Bachay , Ali Abbas Alzaheiree , Hassenien Ali Hussein , Ahmed Nooruldeen Alsafi , Mohammed Hasan Abdulameer
White blood cells (WBCs) are an essential part of the human immune system, playing a significant role in fighting diseases and infections. Their detection and classification from microscopic blood images is a crucial step in diagnosing various diseases. Looking at cells by hand is still key, but it takes a lot of work and mistakes can happen. So, this study tries to improve how to find and sort WBCs using some cool computer tricks. The study tackling issues like cells being on top of each other, looking different, and not having a ton of data. To achieve this, image enhancement techniques were applied using contrast enhancement algorithm, contrast-limited adaptive histogram equalization (CLAHE), and image segmentation techniques using color isolation are employed, which contributes to more accurate separation of overlapping cells, and enables faster and more efficient diagnosis. To efficiently complete the classification process after the segmentation process, a neural network structure consisting of combining three types of convolutional layers (depthwise, spatially, and convolution) was used. To evaluate the proposed technique, experiments were conducted using an open-source blood cell count and detection (BCCD) dataset from the Kaggle platform, and resulted in achieving a classification accuracy of 99.06% and an F1-score of 99.05%. This highlight of the model’s ability to efficiently deal with the challenges associated with WBC classification.
Volume: 15
Issue: 3
Page: 2595-2605
Publish at: 2026-06-01

Deep hybrid models for bitcoin forecasting: EMD, CEEMDAN,and LSTM in comparison

10.11591/ijai.v15.i3.pp2797-2810
Ayoub Aarabi , Maryem Ait Moulay , Issam Bouganssa , Abdelali Lasfar
In this study, an artificial neural network (ANN) was developed to forecast Bitcoin prices using one of the most successful deep learning architectures for time series analysis: long short-term memory (LSTM) networks. This model was enhanced with a signal processing layer that reduces the impact of the instrument’s high volatility on prediction accuracy by applying two signal decomposition techniques: empirical mode decomposition (EMD) and complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN). This study is motivated by the major fluctuations in Bitcoin prices, which make precise forecasting difficult but crucial for experts and investors. This findings demonstrate that forecasting performance improves when decomposition techniques are used. In particular, compared to the conventional LSTM and EMD-LSTM models, the CEEMDAN-LSTM model achieved the highest accuracy, with a mean absolute error (MAE) of 167.837 and a root mean square error (RMSE) of 255.673, outperforming both EMD-LSTM (MAE =168.785, RMSE =256.042) and the standard LSTM (MAE =169.516, RMSE=256.225). The combination of CEEMDAN and LSTM results in a more reliable model that can accurately capture short-term fluctuations in Bitcoin prices.
Volume: 15
Issue: 3
Page: 2797-2810
Publish at: 2026-06-01

Advanced IoT-integrated real-time fire detection and automated mitigation system

10.11591/ijict.v15i2.pp861-868
Rama Krishna Peddarapu , Ajimera Abhinav , Gnana Sathwika V. N. V. , Poosa Brijesh , Amrutha Varshini Ravula
In the field of industry and commerce safety, tackling the most challenging and ongoing fire threats requires the advance internet of things (IoT) integrated real-time fire detection and automated mitigation system. Leveraging IoT and multi-modal sensing in fire safety, the system combines flame, gas, and humidity sensors and cameras to provide continuous real time monitoring and appropriate management of the threats. Real-time automated hazard interventions, such as sprinkler system engagement and geocoded alerts to fire departments, significantly improve life safety outcomes of the system. Active damage mitigation IoT devices provide integrated damage mitigation safety and individual IoT device remote monitoring. In the scope of industry and commerce, this system is a demonstration of the impact of IoT on improving fire safety.
Volume: 15
Issue: 2
Page: 861-868
Publish at: 2026-06-01

Hybrid deep neural network model for aspect and opinion extraction with multi-head attention-driven sentiment analysis

10.11591/ijict.v15i2.pp769-777
Abhinandan Shirahatti , Ramesh Medar , Vijay Rajpurohit , Sanjeev Kaulgud , Mrutyunjaya Mathad Shivamurthaiah
Finding and extracting significant features from review sentences is known as aspect triplet extraction, and it provides succinct information on the elements that users have addressed. This method makes sentiment analysis and opinion mining easier, which helps to provide an adequate understanding of user opinions in reviews. This research presents a novel approach to achieve aspect-sentiment triplet-extraction (ASTE) using a deep neural network and transformer-based multi-head attention model. The proposed hybrid model adopts a pipeline methodology, concurrently extracting opinions and aspects while performing sentiment classification. The study addresses the intricate challenge of identifying triplets that capture nuanced relationships between terms and sentences, employing a deep neural network for joint extraction of aspects and opinions using a sequential tagging method. Sentiment classification is seamlessly integrated into the pipeline, treating sentiment recognition as a classification task, and aspect and opinion extraction as text-extraction challenges. Evaluations was out experimentally on the SemEval 2016 restaurant dataset demonstrate the effectiveness of the model, despite issues with unequal distribution of data.
Volume: 15
Issue: 2
Page: 769-777
Publish at: 2026-06-01

Early prediction of myocardial infarction using proposed score tree algorithm

10.11591/ijict.v15i2.pp813-822
Nusrat Parveen , Utkarsha Pacharaney , Gayatri Hegde , Mohammad Rafique , Sana Firoj Nalband , Shamim Akhtar , Satish Devane
Early detection and diagnosis of a diseases will have a big impact on the medical field and help to prevent loss of life. This study begins by gathering information on myocardial infraction patients from hospitals and focuses on earlier diagnostics. In fact, the pre-processed, confirmed data from a qualified doctor is used for this research. Early prediction of myocardial infarction (MI) is proposed by many researchers. They have used Kaggle datasets that is not recent, and they work on post MI. We have proposed early myocardial infraction detection works on unsupervised datasets. To identify myocardial infraction, numerous machines learning supervised algorithms, including decision tree (DT), random forest (RF), are employed in the literature. In this study, we use the score tree algorithm (STA), which operates on an unsupervised dataset, to present a unique early MI prediction method.
Volume: 15
Issue: 2
Page: 813-822
Publish at: 2026-06-01
1 2 3
...
2091 2092
Show 61 of 2092

Discover Our Library

Embark on a journey through our expansive collection of articles and let curiosity lead your path to innovation.

Explore Now
Library 3D Ilustration