Article Per Year (5 Year)
p-Index From 2021 - 2026
2.986
P-Index
This Author published in this journals
All Journal IAES International Journal of Artificial Intelligence (IJ-AI) Media Statistika Medical Journal of Indonesia Jurnal Penelitian Karet Jurnal Teknologi dan Sistem Komputer Journal of Mathematics and Mathematics Education (JMME) Indonesian Journal of Electrical Engineering and Computer Science Nusantara Science and Technology Proceedings Jambura Journal of Biomathematics (JJBM) Jurnal Ilmiah Matematika Annals of Mathematical Modeling Multidiciplinary Output Research for Actual and International Issue (Morfai Journal) Eduvest - Journal of Universal Studies Annals of Mathematical Modeling Indonesian Journal of Statistics and Its Applications
Bustamam, Alhadi
Departemen Matematika, Fakultas Matematika Dan Ilmu Pengetahuan Alam, Universitas Indonesia
Religion Aerospace Engineering Agriculture, Biological Sciences & Forestry Arts Humanities Chemical Engineering, Chemistry & Bioengineering Civil Engineering, Building, Construction & Architecture Computer Science & IT Decision Sciences, Operations Research & Management Economics, Econometrics & Finance Education Electrical & Electronics Engineering Energy Engineering Environmental Science Health Professions Languange, Linguistic, Communication & Media Law, Crime, Criminology & Criminal Justice Library & Information Science Materials Science & Nanotechnology Mathematics Medicine & Pharmacology Neuroscience Social Sciences Other

Published : 19 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 19 Documents
Search

 1   2 
Arrhythmia Classification Using the Deep Learning Visual Geometry Group (VGG) Model Rudolf Bob Martua B.; Alhadi Bustamam; Hermawan
Nusantara Science and Technology Proceedings Multi-Conference Proceeding Series E
Publisher : Future Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11594/nstp.2023.3702

Abstract

Cardiovascular disease (CVD) is one of the non-communicable diseases (NCDs) and 32% of the world's people die prematurely due to cardiovascular disease (WHO, 2022). The development of computing technology and artificial intelligence (AI), especially Deep Learning (DL), has contributed significantly to helping medical personnel carry out initial pre-diagnosis and classification of heart disease. In this study, we limit heart rhythm detection research into two categories, namely, Normal (N) and Abnormal (An) which are visualized in a standardized amplitude vs time diagram on the PTBDB dataset. The classification model in this research uses the 1-dimensional Deep Neural Network (1D-DNN) Visual Geometry Group, namely, VGG11, VGG13, VGG16, and VGG19. The denoising technique presented in this study on each ECG data sample thereby improving the quality of training data for the AI detection model. The performance of the VGG16 model shows the best training and validation accuracy with the lowest loss, which is 97.85% accuracy; 97.99% precision; 99.75% recall; and 98.52% f1-score. In this way, medical personnel will be helped more quickly in efforts to prevent and control heart disease that occurs in society, especially in the lower middle class. Further research needs to be done to use VGG with more blocks if the structure of the dataset to be classified is much more complex.
Reconstruction of the Phi-2 Method for Question-Answering Related to Diabetes Disease Using the MedAlpaca Dataset Ridho, Muhammad; Bustamam, Alhadi; Adnan, Risman
Jambura Journal of Biomathematics (JJBM) Volume 6, Issue 3: September 2025
Publisher : Department of Mathematics, Universitas Negeri Gorontalo

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.37905/jjbm.v6i3.30506

Abstract

This  study  focuses on the reconstruction of the Phi-2  method  for text-based question-answering systems  related to diabetes  using the MedAlpaca dataset.   The  aim  is to enhance  the accuracy in  diabetes  question-answering applications.   We  leverage LoRA  techniques   to fine-tune  the model,  thereby  improving its  ability to handle complex medical queries.  The integration of the MedAlpaca dataset, which contains  a diverse range of medical questions  and answers,  provides a robust  foundation for training and testing the model.  The results  reveal  that fine-tuning  with   MedAlpaca  significantly  enhances   the  model’s   performance,  achieving  higher   accuracy compared to the base Phi-2  model,  achieving a performance increase  from  14.81% to 49.37% on MedMCQA, reaching  92.83%  on  PubMedQA, and  38.78%  on  MedQA. It  also  surpasses  other  leading  models   such  as BioBERT  (89.90%)   and   GatorTron  (90.87%).        The   results    highlight  the   effectiveness    of   incorporating domain-specific datasets  like  MedAlpaca to boost model  performance.  This  advancement points  to promising directions  for  future  research,   including  expanding datasets  and  refining fine-tuning techniques   to  further improve automated  medical question-answering systems.
COMPARISON OF MISSING VALUE IMPUTATION USING MEAN, BAYESIAN KNN, AND NON-BAYESIAN KNN ON TEP GENE EXPRESSION DATA Mastika, Mastika; Siswantining, Titin; Bustamam, Alhadi
MEDIA STATISTIKA Vol 18, No 1 (2025): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/medstat.18.1.61-72

Abstract

Analysis of gene expression data, particularly in cancer data, often faces challenges due to the presence of missing values. One approach to overcome this is data imputation. This study evaluates the performance of three imputation methods, namely mean imputation, K-Nearest Neighbors (KNN), and KNN with Bayesian optimization using Gaussian Process modeling, on Tumor Educated Platelets (TEP) gene expression data. Missing values were introduced using Missing Completely at Random (MCAR) gradually at levels of 5%, 10%, 15%, and up to 60%, and performance was evaluated using three metrics: Mean Absolute Error (MAE), Mean Squared Error (MSE), and Normalized Root Mean Squared Error (NRMSE). The results show that the three methods produce relatively similar performance, with differences in MAE, MSE, and NRMSE values only at a small decimal scale. Although Bayesian Optimization is expected to improve the accuracy of KNN, the resulting improvement on this dataset is not significant. These findings indicate that simple imputation such as the average and KNN-based methods still provide competitive results on TEP data with data characteristics that have 14,020,496 zeros out of a total of 16,512,496 existing values, which is approximately 84.91% of the total data.
Analysis of diabetes mellitus gene expression data using two-phase biclustering method Kafi, Rahmat Al; Bustamam, Alhadi; Mangunwardoyo, Wibowo
Jurnal Ilmiah Matematika Vol 8, No 2 (2021)
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/konvergensi.v0i0.22111

Abstract

The purpose of this research is to find bicluster from Type 2 Diabetes Mellitus genes expression data which samples are obese and lean people using two-phase biclustering. The first step is to use Singular Value Decomposition to decompose matrix gene expression data into gene and condition based matrices. The second step is to use K-means to cluster gene and condition based matrices, forming several clusters from each matrix. Furthermore, the silhouette method is applied to determine the number of optimum clusters and measure the accuracy of grouping results. Based on the experimental results, Type 2 Diabetes Mellitus dataset with 668 selected genes produced optimal biclusters, with six biclusters. The obtained biclusters consist of 2 clusters on the gene-based matrix and 3 clusters on the sample-based matrix with silhouette values, respectively, are 0.7361615 and 0.7050163.
CONV1D-LSTM-BASED QSAR CLASSIFICATION MODEL FOR BACE1 INHIBITORS: A COMPREHENSIVE APPROACH WITH DESALTING, PAINS FILTERING AND DRUG-LIKENESS ANALYSIS Trianto Haryo Nugroho; Alhadi Bustamam
Multidiciplinary Output Research For Actual and International Issue (MORFAI) Vol. 5 No. 3 (2025): Multidiciplinary Output Research For Actual and International Issue
Publisher : RADJA PUBLIKA

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54443/morfai.v5i3.3023

Abstract

In recent years, the discovery of Beta-Secretase 1 (BACE1) enzyme inhibitors for more effective Alzheimer’s therapy has become a major focus, making in silico research to identify new inhibitors with minimal side effects increasingly essential. Ligand-Based Virtual Screening (LBVS) using Quantitative Structure–Activity Relationship (QSAR) methods offers a fast and cost-effective alternative to experimental assays. In this study, we propose a Conv1D-LSTM-based QSAR model as a novel approach for classifying BACE1 enzyme inhibitors, where Conv1D is employed for encoding molecular data and LSTM is used to classify compounds as active or inactive. The model is complemented by drug-likeness analysis based on Lipinski's Rule of Five to evaluate the therapeutic potential of candidate molecules. The dataset used includes 711 molecular structures, consisting of 278 active and 433 inactive compounds. Experimental results demonstrate that our model achieves a classification accuracy of 79.13%, with a sensitivity of 73.02%, specificity of 83.08%, and a Matthews Correlation Coefficient (MCC) of 56.38%.
Versatile, low-cost ophthalmic wet lab device to improve diagnostic and surgical eye training Mardianto, Umar; Victor, Andi Arus; Yusuf, Prasandhya Astagiri; Juniantito, Vetnizah; Kekalih, Aria; Rahayu, Tri; Bustamam, Alhadi; Edwar, Lukman
Medical Journal of Indonesia Vol. 35 No. 1 (2026): March
Publisher : Faculty of Medicine Universitas Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.13181/mji.bc.257865

Abstract

Ophthalmologists rely on wet lab training for both diagnostic procedures and surgical techniques. Existing wet lab devices are limited to surgical training and lack functionality for performing required perioperative diagnostic examinations. This study aimed to develop an affordable, easily manufactured eye holder to enhance ophthalmology training for wet lab simulations. A three-dimensional (3D)-printed animal eye holder was designed in 3D with a funnel-shaped structure resembling an orbital eye socket. The design was optimized for optimal wet lab activities. The animal eye holder device demonstrated potential use for ultrasound biometry, handheld keratometry, tonometry, and ophthalmological surgical training. These activities can be performed effectively after the animals’ eyes are stabilized inside the holder in flat and inclined positions. This innovative animal eye holder is the first designed to provide flexible diagnostic practice and surgical training, especially during wet lab activities.
MODEL KLASTERING SKM3 (SUBCONTROLLED K-MEANS MAX-MIN) DAN APLIKASINYA DALAM MENGHITUNG ELEKTABILITAS PASANGAN CALON KEPALA DAERAH Tampubolon, Patuan P; Kaloka, Tesdiq Prigel; Swasti, Olivia; Mustika, Widya Fajar; Bustamam, Alhadi
Journal of Mathematics and Mathematics Education Vol 8, No 2 (2018): Journal of Mathematics and Mathematics Education (JMME)
Publisher : Universitas Sebelas Maret

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20961/jmme.v8i2.25838

Abstract

Abstract: Indonesia is a legal state that chooses a leader based on the results of general elections, such as the election of presidents and regional leaders. Electability is statistical data for each pair of candidates who show public interest to choose the candidate. Electability data is usually obtained from the results of questionnaires or interviews with constituents. The data search process is carried out by a survey institution. Most people discuss voluntarily in social media related to the candidate that they will choose. This study uses discussion data from social media to calculate the electability of each pair of candidates by using cluster method. The cluster method is K-Means. K-Means employs euclidean distance to determine the cluster of each data, while the number of cluster can be determined by the user. This study proposes SKM3 model (Subcontrolled K-Means Max-Min), which applies the minimum and maximum average values to decide the cluster of each data. SKM3 cluster is controlled by K-Means method that uses Euclidian distance. SKM3 model is processed using news data from detik.com site for the election of regional leader of West Java, Central Java, and East Java. The error value of SKM3 model is calculated through RMSE (Root Mean Square Error). The error value of West Java is 0.0452, the error value of Central Java up to 0.0343, and the error value of East Java is 0.2382. Based on the error values of each electoral region, it shows that SKM3 model has a small error value, so it can be concluded that SKM3 model is good for calculating the electability of the leader by using clustering method.Keywords:Electability, Clustering, K-Means, SKM3.
A Comparative Study of Sequential Biclustering and Fuzzy C-Means, K-Nearest Neighbors, and Mean Imputation for Missing Value Estimation in Gene Expression Data Yolanda Azzahra; Titin Siswantining; Setia Pramana; Mogana Darshini Ganggayah; Alhadi Bustamam
Indonesian Journal of Statistics and Applications Vol 10 No 1 (2026): Vol 10 Issue 1 June 2026
Publisher : Statistics and Data Science Program Study, SSMI, IPB University, in collaboration with the Forum Pendidikan Tinggi Statistika Indonesia (FORSTAT) and the Ikatan Statistisi Indonesia (ISI)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29244/ijsa.v10i1p37-46

Abstract

Missing values are a common issue in gene expression data and can significantly affect downstream analysis. This study aims to compare the performance of a hybrid sequential biclustering and centroid-based clustering method with conventional imputation approaches for handling missing values. The proposed method integrates sequential biclustering based on mean squared residue to identify coherent submatrices, followed by centroid-based clustering to estimate missing entries. The dataset used in this study consists of gene expression data of patients with type 2 diabetes mellitus, with missing values introduced under various proportions ranging from 5% to 55%. The performance of the proposed method is evaluated and compared with mean imputation and nearest neighbor imputation using mean squared error, root mean squared error, and mean absolute error. The experimental results show that the proposed method consistently produces lower errors across all missing rates than the baseline methods. This indicates that incorporating local pattern structures through biclustering improves the accuracy of missing value estimation. The findings suggest that the proposed hybrid framework is more effective in preserving the underlying structure of gene expression data and provides a reliable approach for handling missing data in high-dimensional biological datasets.
Evaluation And Selection Of Optimal Deep Learning Architecture For Predicting The Endpoint In High Shear Wet Granulation For Antacid Tablet Production Irvan Maulana; Arry Yanuar; Sutriyo Sutriyo; Alhadi Bustamam
Eduvest - Journal of Universal Studies Vol. 4 No. 6 (2024): Journal Eduvest - Journal of Universal Studies
Publisher : Green Publisher Indonesia

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.59188/eduvest.v4i6.1274

Abstract

Objective: The purpose of this research was to evaluate and select the best architecture among native convolutional neural network (CNN), MobileNetV2, ResNet50V2, and EfficientNetB0 for predicting the endpoint of the high shear wet granulation process, with accuracy as the main evaluation metric. Methods: The dataset was captured from an industrial camera using static image analysis and was manually labeled as “NOT READY” and “READY” according to the traditional endpoint method based on the mixer’s ampere point in the granulator. The dataset contained a total of 180 images, which were split between training and validation sets. Native CNN and TensorFlow Keras application programming interface (API) were utilized with MobileNetV2, EfficientNetB0, and ResNet50V2 as base feature encoders. Hyperparameters, such as final Fully Connected (FC) layer width, dropout rate, and learning rate, were optimized for binary classification using Keras hyper tuning. Results: The best was the native CNN, it was also the fastest among the three other models, taking only 20-30 ms per step for inference during runtime, though it requires 9000 ms time for training, the longest time among the models. It achieved an accuracy of 98%, and a validation accuracy of 97%. Conclusion: The system was able to determine when a wet granulation process has reached its endpoint based on live images from a camera after being trained on previously labeled data. The native CNN was the best model, offering the fastest runtime performance and the highest accuracy.
Co-Authors AA Sudharmawan, AA Adnan, Risman Amnia Salma Andi Arus Victor Aria Kekalih Arry Yanuar Ayu, Farida Devvi Sarwinda Devvi Sarwinda Dian Maharani Dwi Guna Mandhasiya Efriana, Anisya Feby Harlianto, Didi Hendri Murfi Hermawan Irvan Maulana Kafi, Rahmat Al Kartika Chandra Dewi Lukman Edwar Mardianto, Umar MASITA DWI MANDINI MANESSA, MASITA DWI MANDINI Mastika, Mastika Mogana Darshini Ganggayah Mohamad Irlin Sunggawa Muhammad Ridho Mustika, Widya Fajar Nadya Asanul Husna NadyaAsanul Husna Nyoman A. Wibawa Nyoman A. Wibawa Prasandhya Astagiri Yusuf Prasnurzaki Anki Rachardi, Andris Rahmat Al Kafi Rahmat Al Kafi Retno Lestari Riesnandar, Ariq Anggaraksa Rudolf Bob Martua B. Rusdah, Deandra Aulia Safitri, Egi Sarwinda, Devvi Selly A. A. K Selly Anastassia Amellia Kharis Setia Pramana Siswantining, Titin Stevanuse, Charlos Togi SUDARSONO, ELY Supriatna Supriatna Sutriyo Sutriyo Swasti, Olivia Tampubolon, Patuan P Tesdiq Prigel Kaloka Titin Siswantining Tri Rahayu Trianto Haryo Nugroho Vetnizah Juniantito WIBOWO MANGUNWARDOYO Wibowo Mangunwardoyo Yolanda Azzahra Yuni Rosita Dewi Zuherman Rustam