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

30,938 Article Results

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

Forecasting world sugar contract futures using long short-term memory technique with multi-step ahead forecasting strategy

10.11591/ijai.v15.i3.pp2633-2642
Khairil Anwar Notodiputro , Kayla Fakhriyya Jasmine , Indahwati Indahwati , Wandee Wanishsakpong
Time series analysis using stochastic and dynamic models for data forecasting is a key in assisting planning and decision-making processes in various sectors. Long short-term memory (LSTM), with its advantage in understanding patterns and non-linearity in sequential data, is applied in a multi-step ahead forecasting strategy on world sugar futures prices. Fluctuations in sugar prices have a significant impact on the agriculture, trade, and food industry sectors. Forecasting sugar prices becomes a crucial tool for industries, investors, and traders to anticipate changes and make informed decisions. The objectives of this study are to identify the best strategy for forecasting the world sugar contract price and to perform forecasting using the best model. The research results indicate that hyperparameter tuning in LSTM models produces varied combinations and effects. Furthermore, the recursive strategy is suitable for long-term forecasting, while the direct strategy is appropriate for short-term forecasting. Forecasting values for long-term periods remains challenging in achieving high accuracy.
Volume: 15
Issue: 3
Page: 2633-2642
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

From audio to image: gunshot classification using Mel spectrogram convolutional neural networks

10.11591/ijai.v15.i3.pp2166-2180
Peerapol Khunarsa , Pafan Doungpaisan
Accurate identification of firearm types from acoustic signals is essential for modern public safety and forensic applications. Traditional gunshot analysis methods often rely on physical evidence or handcrafted audio features, which can be unreliable under noisy and reverberant conditions. This study presents a systematic investigation of gunshot sound classification using Mel spectrogram representations and convolutional neural networks (CNNs). Raw audio signals are transformed into Mel spectrogram images, enabling firearm classification to be formulated as an image recognition problem. Thirteen CNN architectures, ranging from lightweight to deep models, are evaluated under a unified experimental protocol to analyze both classification performance and computational efficiency. Experiments are conducted on a publicly available multi-firearm dataset recorded in semi-controlled real-world environments. The results demonstrate that Mel spectrogram–based CNN models achieve classification accuracy exceeding 94%, while moderate-complexity architectures provide a favorable balance between accuracy and efficiency. The findings highlight the importance of representation–architecture alignment and offer practical design guidelines for selecting deployable CNN models in real-time gunshot detection systems.
Volume: 15
Issue: 3
Page: 2166-2180
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

Student satisfaction in student affairs management: the role of cross-functional cooperation in Hainan, China

10.11591/ijere.v15i3.39033
Erlin Tian , Supot Rattanapun
Student affairs management (SAM) is increasingly expected to deliver timely, coherent, and student-centered services, yet satisfaction remains uneven in Chinese universities because students experience SAM as an integrated system rather than isolated units. This study asks whether cross-functional cooperation (CFC) explains how service quality, service gaps, and students’ psychological and engagement factors translate into satisfaction with SAM in Hainan Province, China. Using a cross-sectional survey of 250 undergraduate and postgraduate students from ten public and private universities, the study applies partial least squares–structural equation modeling (PLS-SEM) with 5,000-sample bootstrapping to test direct and mediating effects. Results show that CFC is the strongest predictor of satisfaction (β=0.419, p<0.001). Service gaps reduce satisfaction (β=−0.217, p<0.001), while psychological and engagement factors increase satisfaction (β=0.206, p<0.001). Service quality has no direct effect but operates through CFC, indicating that coordination is required to convert service inputs into positive experiences. The findings highlight governance reforms that institutionalize cross-department coordination, shared case management, and gap monitoring to improve SAM effectiveness under Hainan’s reform context.
Volume: 15
Issue: 3
Page: 1963-1970
Publish at: 2026-06-01

Threat appraisal and prevention of risky sexual behavior among high school students in Indonesian: the mediating roles of response efficacy and self-efficacy

10.11591/ijere.v15i3.38765
Erni Gustina , Ira Nurmala , Nunik Puspitasari
Premarital sexual behavior among adolescents remains a public health concern. However, school programs often focus on risk perception without showing how perceived threat can lead to protection. This study examined the influence of threat appraisal (severity and vulnerability) on the prevention of premarital sexual behavior through response efficacy and self-efficacy as mediators. A cross-sectional survey was conducted among 333 high school students selected by multistage sampling. Likert-scale measures assessed severity, vulnerability, response efficacy, and self-efficacy. The model was analyzed using partial least squares structural equation modeling (PLS-SEM) with bootstrapping (α=0.05). Most participants reported no premarital sexual behavior (66.1%). Severity and vulnerability positively predicted response efficacy and self-efficacy, with severity exerting a stronger effect. Response efficacy and self-efficacy were negatively linked to premarital sexual behavior (p<0.05), meaning that higher coping appraisal (the ability to handle risky situations) was associated to lower risk behavior. Indirect effects from severity and vulnerability to behavior through both mediators were significant. The model explained 34% of the variance in premarital sexual behavior. Threat appraisal reduced premarital sexual behavior mostly by improving coping appraisal. Thus, school-based interventions should combine risk communication with the development of refusal, negotiation, and self-confidence skills to support protection.
Volume: 15
Issue: 3
Page: 2390-2398
Publish at: 2026-06-01

ChatGPT in the university classroom: perceptions, perceived usefulness and use intentions among undergraduate students

10.11591/ijere.v15i3.38709
Vidnay Noel Valero-Ancco , Yolanda Lujano-Ortega
The objectives of the present study are to analyze university students’ perceptions of and attitudes toward ChatGPT as a support tool for learning, as well as the user profiles derived from their technological acceptance, were analyzed. A quantitative, nonexperimental and cross-sectional design was applied to a non-probabilistic convenience sample of 438 undergraduate students, and a validated questionnaire composed of three dimensions compatibility with learning styles, ease of use and perceived usefulness, and continued use intentions was used. The data were analyzed with hierarchical cluster analysis (Ward method). The results reveal three groups of users: i) enthusiasts, with highly favorable perceptions and high use intentions; ii) moderate users, who exhibit partial acceptance; and iii) critical or disconnected users, who have a negative view of the tool. Taken together, the findings confirm that perceived usefulness, ease of use and trust are determining factors in the intention to use ChatGPT. This study provides empirical evidence on the diversity of attitudes toward generative artificial intelligence (GenAI) in university contexts and highlights the need to promote institutional strategies of critical digital literacy and teacher training for its ethical and pedagogical integration.
Volume: 15
Issue: 3
Page: 2073-2081
Publish at: 2026-06-01

Narrative comprehension in 5-year-old Vietnamese-speaking children using the multilingual assessment instrument for narratives

10.11591/ijere.v15i3.38932
Nguyen Thi Hoang Yen , Ben Phạm , Hang Pham , Van Pham , Phuong Nguyen , Phuong Bui
This study addresses how elicitation modes and socio-demographic factors influence narrative macrostructure understanding in 5-year-old Vietnamese-speaking children. A convenience sampling of 311 typically developing children were assessed using the multilingual assessment instrument for narratives (MAIN). Narrative comprehension was evaluated through 10 standardized questions for both retelling (cat story) and storytelling (baby goats story) modes. Findings revealed a significant advantage for retelling (M=7.21, SD=2.31) over storytelling (M=5.73, SD=2.50; p<.001), highlighting the role of linguistic scaffolding. While children mastered identifying character goals, challenges remained in explaining internal states and making causal inferences. Narrative comprehension was independent of gender, location, and general communication skills, but significantly influenced by maternal education level (p<.05). Vietnamese narrative development follows universal patterns, yet deep comprehension is shaped by specific environmental inputs. This study establishes a normative baseline for narrative skills within the Vietnamese preschool curriculum and provides a validated tool for speech and language therapists to facilitate early identification and targeted interventions.
Volume: 15
Issue: 3
Page: 1908-1918
Publish at: 2026-06-01

Instructional scaffolding in dialogue-based programming tutoring

10.11591/ijere.v15i3.38919
Julieto Perez , January Naga , Salma Naga-Marohombsar
This study examines how instructional scaffolding is enacted in dialogue-based artificial intelligence (AI) tutoring systems for programming education and evaluates the levels of cognitive demand they support. While AI tutors can guide novice learners through programming tasks, it remains unclear whether they promote meaningful higher-order thinking or primarily support procedural task completion. Using a mixed-methods approach, 1,255 tutor utterances from 36 tutoring sessions were analyzed using a dual-layer coding framework grounded in instructional scaffolding theory and Bloom’s revised taxonomy. Results show that instructional support is concentrated at the understanding and applying levels, with prompting and explaining as dominant strategies. Higher-order cognitive scaffolding (analyzing, evaluating, creating) was rare or absent. Sequential patterns revealed repetitive prompting–explaining cycles with limited scaffold progression. These findings indicate that AI tutoring effectively supports foundational learning but lacks mechanisms for deeper cognitive engagement. This study highlights the need for pedagogically informed AI tutor design and provides actionable insights for educators and system developers to integrate AI tools in ways that promote higher-order thinking and independent problem-solving.
Volume: 15
Issue: 3
Page: 2478-2486
Publish at: 2026-06-01

Beyond diagnosis: using PNImodified and composite priority indices to orchestrate meta-skills-driven academic management innovation

10.11591/ijere.v15i3.38478
Chi Che , Sukanya Chaemchoy , Pruet Siribanpitak
This study translates an academic management–meta-skills integration framework into a data-driven innovation roadmap for private higher education institutions (HEIs) in Sichuan, China. Using an explanatory sequential mixed-methods design, Phase 1 surveyed 400 undergraduates who provided dual ratings of current performance (degree of success, D) and desired priority (importance, I), enabling computation of the modified priority needs index (PNImodified=(I−D)/D) across meta-skills domains and academic management subcomponents. In Phase 2, institutional leaders and senior academics rated the feasibility and impact of aligned innovations; these ratings were integrated with PNImodified to calculate a composite priority index (CPI) and propose phased implementation sequencing. Results indicated the largest perceived meta-skills development needs in adaptive expertise (PNImodified=0.48) and relational dynamics (0.34). At the academic management level, curriculum development (0.56) and evaluation and assessment (0.45) emerged as the most critical domains. Curriculum structuring (0.65), instructional design (0.53), meta-skills evaluation modules (0.59), and learning engagement (0.59) consistently ranked as top subcomponent priorities and were positioned as Phase 1 actions in CPI-based sequencing. Experts rated the overall innovation as highly suitable (M=4.55) and feasible (M=4.53). The combined indices provide a practical decision tool for sequencing meta-skills-oriented academic management innovations in Sichuan private HEIs.
Volume: 15
Issue: 3
Page: 1862-1875
Publish at: 2026-06-01

Pakistan English language policy alignment with IDLE-informed policy model

10.11591/ijere.v15i3.38739
Waqas Ahmad , Muhammad Taufiq Al Makmun
English controls academic and professional access in Pakistan, yet the National Education Policy Development Framework (NEPDF) 2024 completely ignores informal digital learning of English (IDLE), which refers to self-directed learning through digital tools: WhatsApp, YouTube, and chatbots. No prior study has examined this policy-practice gap within Pakistan’s post-2024 framework, particularly across urban and rural communities in Khyber Pakhtunkhwa, Punjab, and Sindh. This qualitative case study gathered perspectives from 20 undergraduate students, 10 teachers, and 5 policymakers through semi-structured interviews, focus groups, and content analysis of NEPDF 2024 and provincial policy texts, analyzed using NVivo-facilitated STAP thematic analysis. Findings show that students and teachers actively use IDLE tools while policymakers remain largely unaware. The IDLE-informed policy model (IIPM), grounded in connectivism, sociocultural theory (SCT), and learner autonomy, is proposed as a practical policy framework that incorporates low-bandwidth tools like WhatsApp to expand access for under-resourced learners. This study contributes to educational evaluation by assessing the alignment between Pakistan’s national language policy and grassroots IDLE practices, producing a transferable policy evaluation model for global south English as a foreign language (EFL) context.
Volume: 15
Issue: 3
Page: 2194-2204
Publish at: 2026-06-01

Residual reinforcement learning for disturbance-resilient control under modeling uncertainties

10.11591/ijece.v16i3.pp1175-1187
Abolanle Adetifa , Rexcharles Enyinna Donatus , Daniel Udekwe
Modern control systems must operate reliably in the presence of modeling uncertainties and external disturbances, conditions under which conventional fixed-gain controllers often exhibit performance degradation. This paper proposes a residual reinforcement learning framework for disturbance-resilient pitch-rate control of an aircraft longitudinal model. A classical proportional-integral-derivative (PID) controller is employed as a stabilizing baseline, while a deep deterministic policy gradient (DDPG) agent learns a bounded residual control signal to compensate for unmodeled dynamics and external perturbations. To promote favorable transient behavior, the learning process incorporates transient-aware and reference-model-based reward shaping, while actuator constraints are enforced within the environment dynamics. Simulation results demonstrate that the proposed residual controller achieves a superior balance between response speed, overshoot, and tracking accuracy compared with both the standalone PID controller and a pure DDPG-based controller. In particular, the residual architecture significantly reduces overshoot and tracking error while preserving fast transient response and providing robust disturbance rejection under large pitching moment disturbances. These results indicate that residual reinforcement learning offers a practical and effective approach for enhancing robustness and performance in safety-critical flight control applications.
Volume: 16
Issue: 3
Page: 1175-1187
Publish at: 2026-06-01

Effectiveness of HIPO Android-based application in improving hypertension self-management literacy among obese patients

10.11591/ijphs.v15i2.27015
Ros Endah Happy Patriyani , Sunarsih Rahayu
Hypertension in obese patients requires comprehensive management through enhanced health literacy. Android-based applications represent a promising innovation for improving hypertension control literacy. To analyze the effectiveness of the Android-based hypertension and obesity information system (HIPO) application in improving hypertension control literacy among obese patients. This quasi-experimental study employed a one-group pretest-posttest design without a control group, conducted at Sibela Community Health Center, Surakarta, from June to December 2025. Seventy-six respondents were selected through purposive sampling, meeting criteria of primary hypertension (≥ 140/90 mmHg) and BMI ≥ 25 kg/m². The intervention involved four weeks of HIPO application use. Data were analyzed using the Wilcoxon Signed Rank Test. Respondents were predominantly female (67.11%), aged 46-55 years (31.58%), with a genetic predisposition to hypertension (57.89%). Systolic blood pressure significantly decreased from 159.53 ± 14.19 to 144.48 ± 11.44 mmHg (p < 0.001), and diastolic from 95.20 ± 6.55 to 88.50 ± 4.58 mmHg (p < 0.001). Good knowledge increased from 35.53% to 55.26%, and good hypertension control increased from 35.53% to 61.84%. The HIPO application significantly improved blood pressure, knowledge, and hypertension control. However, the absence of a control group limits causal inference. This application may serve as a supplementary educational tool in primary healthcare chronic disease management programs.
Volume: 15
Issue: 2
Page: 408-416
Publish at: 2026-06-01

Vietnamese secondary school EFL students’ motivation and attitudes toward using ChatGPT for vocabulary learning

10.11591/ijere.v15i3.38603
Thuan Anh Pham , Cuong Huy Pham
The escalating adoption of artificial intelligence (AI) in language education, particularly ChatGPT, has drastically altered vocabulary teaching and learning practices in secondary schools in Vietnam, where rote learning and teacher-centered approaches are common norms. This study examines secondary school students’ motivation and attitudes toward using ChatGPT for vocabulary development. A mixed-methods design was employed, drawing on a survey with 208 ninth-grade students and semi-structured interviews with eight of the respondents. The survey data revealed the students’ moderate degrees of motivation to use ChatGPT, with extrinsic motivation surpassing intrinsic motivation. Their attitudes were generally positive, showing relatively high enjoyment and low anxiety in utilizing ChatGPT for vocabulary learning. The students considered ChatGPT a useful and convenient resource and expressed willingness to continue using it despite their concerns over its reliability and the risk of over-reliance. The interview data augmented these patterns, highlighting ChatGPT’s interactivity and flexibility while emphasizing the continued importance of teachers’ classroom support. This study provides corroborating evidence for ChatGPT as a versatile complementary tool for vocabulary learning that fosters learners’ motivation and positive attitudes. However, its adoption requires careful integration with teacher guidance and AI literacy training.
Volume: 15
Issue: 3
Page: 2708-2715
Publish at: 2026-06-01
1 2 3
...
2062 2063
Show 54 of 2063

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