BABilqis Amaliah

Bilqis Amaliah

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Bilqis Amaliah has a total of 55 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A Prescriptive Analytics Framework for Optimizing the Indonesian Salt Supply Chain Using System Dynamics and Data-Driven Decision Support

    by Titus Kristanto, Ahmad Saikhu, Bilqis Amaliah

    The Indonesian salt supply chain is characterized by persistent price volatility, supply-demand imbalances, and limited decision-support tools for policy development. Current models are mostly descriptive or predictive and provide forecasts with little actionable advice. This study presents a prescriptive analytics framework that couples a calibrated system dynamics model with a mixed-integer linear programming optimization model to assist in monthly decision-making regarding buffer stock release, import quota allocation, and distribution routing. The system dynamics model was calibrated using historical data (2020–2025), and the key stocks, flows, and feedback loops in the Indonesian salt supply chain were captured. The optimization model minimizes the weighted objective function of price volatility, service level, and logistics cost under inventory balance, capacity, and non-negativity constraints. An evaluation with 100 stochastic scenarios shows that the proposed framework reduces the average monthly price volatility by 23.4% and improves the service level from 81.6% to 94.1% relative to the baseline heuristic policies, with a 6.3% increase in logistics cost. A sensitivity analysis indicates that a buffer stock capacity of at least 15% of the annual consumption is required to maintain price stability. The proposed framework offers a data-driven decision-support approach for translating supply chain forecasts into actionable policy recommendations.

  • Open Access

    Cardiovascular Disease Classification Using K-Nearest Neighbors with Multiple Imputation by Chained Equations (KNN-MICE)

    by Diana Purwitasari, Nursalam, Arina Qona'AH, Bilqis Amaliah, Dini Adni Navastara, Tegar Ganang Satrio Priambodo

    Cardiovascular disease (CVD) stands as the most prevalent comorbidity and a primary cause of death among diabetes patients. Existing classification approaches often struggle with severe class imbalance in the utilized datasets and missing clinical data, resulting in biased predictions that fail to capture at-risk minority cases. Predicting CVD risk using the Behavioral Risk Factor Surveillance System (BRFSS) dataset is challenging due to its high-dimensional features, missing data, and significant class imbalance. Unlike mean or median imputation, which can distort data distributions and ignore relationships between variables, K-Nearest Neighbors with Multiple Imputation by Chained Equations (KNN-MICE) preserves complex multivariate correlations inherent in clinical datasets. MICE have the ability to navigate expansive feature spaces without relying on the strict parametric assumptions that cause simpler methods to fail. This research introduces an integrated machine learning pipeline that merges KNN-MICE and hybrid sampling to enhance data balance. We aim to train and optimize CVD risk classification models capable of identifying at-risk minority cases without compromising the overall reliability. The proposed model reduced majority-class bias, achieving a Macro-F1 score of 66.87%, Precision of 75.12%, and Recall of 78.22%. Furthermore, SHapley Additive exPlanations (SHAP)-based feature importance analysis identified demographic and intrinsic health factors, particularly age, perceived general health, and sex, as the most significant clinical predictors of CVD risk. This integrated pipeline enhances early CVD detection in diabetic patients while offering valuable clinical insights into key risk factors.

  • Open Access

    Supply Average Method for Finding the Initial Basic Feasible Solution of Transportation Problem

    by Bilqis Amaliah, Ilham Gurat Adillion, Bintang Nuralamsyah, Moch. Nafkhan Alzamzami, Danno Denis Dhaifullah, Ahmad Hoirul Basori, Rihan Farih Bunyamin

    The Transportation Problem (TP) aims to minimize the total cost of distributing items from supplies to demands, with the Initial Basic Feasible Solution (IBFS) as its critical first step. Since existing IBFS methods often fail to deliver high-quality solutions, this research proposes the Supply Average Method (SAM) as an improved method. SAM compares the Average Supply (AS) value, which is the product of each row's average transportation cost and its supply and applies tie breaking rules for First and Second Least Cost cells during allocation. SAM was evaluated on 47 datasets and benchmarked against five established methods (LCM, VAM, TOCM-MT, BCE, and SSM) using Accuracy Percentage, Deviation Percentage (Dv), Improvement Percentage (IP), The Frequency of Numbers of Optimal Cost Solutions (FNOCS), and Number of Iterations. SAM achieved the highest accuracy rate at 82.98%, the lowest Dv average at 0.51%, produced the most positive IP values (indicating a lower IBFS than compared methods), outperformed in four of ten TP matrix sizes, and required the fewest total iterations (275). SAM introduces a cost average based allocation strategy with explicit tie breaking rules absent in prior methods, yielding superior IBFS quality and faster convergence to optimality.

  • Open Access

    FORESIGHT: Frequency and Order-Aware REpresentation With Spectral-Domain and Inter-Channel Global Horizon Transformer

    by Mustafa Kamal, Ary Mazharuddin Shiddiqi, Saad Ali Alahmari, Ervin Nurhayati, Bilqis Amaliah, Bimantara Tito Wahyudi, Arif Bramantoro, Abdulaziz AlMohimeed

    Forecasting multivariate time series remains challenging due to the coexistence of long-term dependencies, periodic structures, and complex inter-variable interactions. Existing methods often emphasize either frequency-aware or order-aware representations, without fully leveraging their complementary roles. To address this, we propose FORESIGHT (Frequency- and Order-aware Representation with Spectral-domain and Inter-channel Global Horizon Transformer), a dual-stream framework that models seasonal and trend components through coordinated representation learning. The model employs an Exponential Order-aware Adaptive (EOA) decomposition to separate input signals, followed by a spectral-channel encoding module (SCTEM) for capturing periodic patterns and inter-variable dependencies, and an order-aware module (ORM) for modeling temporal progression in trend components. An adaptive fusion mechanism integrates both streams in an end-to-end manner. Extensive experiments on five benchmark datasets (Weather, Exchange Rate, Traffic, ETTh1, and ETTm2) across multiple prediction horizons demonstrate competitive and consistent performance relative to strong baselines. Compared with FAITH, FORESIGHT achieves average improvements of 3.092% in MAE and 2.463% in RMSE while maintaining statistical parity under strict significance thresholds. These results suggest that integrating frequency- and order-aware representations leads to more stable and robust long-horizon forecasting.

  • Open Access

    Enhanced supply selection for better initial basic feasible solutions in transportation problems

    by Rihan Farih Bunyamin, Bilqis Amaliah, Ahmad Saikhu

  • Open Access

    Trust Region Plug-and-Play With Provable Convergence for Image Restoration

    by Christian Budhi Sabdana, Dwi Sunaryono, Bilqis Amaliah, Shoffi Izza Sabilla, Laras Suciningtyas

  • Open Access

    Peritoneal Carcinomatosis Detection Based on Improved YOLOv8 With Attention Mechanism and Optimized Backbone Integration

    by Naim Rochmawati, Chastine Fatichah, Bilqis Amaliah, Agus Budi Raharjo, Frédéric Dumont, Emilie Thibaudeau, Cédric Dumas

    Peritoneal Carcinomatosis (PC) represents a severe form of metastatic cancer, with lesions spreading across the peritoneum and often leading to a poor prognosis. Early detection of PC lesions is essential for timely treatment, yet identifying smaller lesions in medical images remains challenging due to their subtle appearance and varied sizes. Traditional detection techniques often fall short in this area, underscoring the need for more advanced solutions. This study introduces improvements to the YOLOv8 architecture aimed at enhancing the detection of small PC lesions in laparoscopic images. The proposed modifications include adding a Convolutional Block Attention Module (CBAM) to improve feature extraction, incorporating a P2 layer for better detection of small objects, and using a CIoU loss function for more effective training. CIoU surpassed DIoU, WIoU, GIoU, EIoU, and SIoU loss functions in performance. Experimental results across Nano, Small, and Medium model scales demonstrated significant gains in Precision, Recall, mAP50, and F1 Score over the original YOLOv8. The modified model achieved a Precision of 0.887, Recall of 0.769, mAP50 of 0.861, and an F1 Score of 0.824. Ablation studies validated the benefits of each modification, with the enhanced model outperforming other YOLO versions (v5–v10) and transformer-based detectors such as RT-DETR, as well as alternative attention mechanisms including ECA and SA. These findings highlight the effectiveness of the proposed YOLOv8 modifications for detecting small lesions in medical imaging.


  • Open Access

    Deep Learning-Based Lesion Detection in Endoscopy: A Systematic Literature Review

    by Naim Rochmawati, Chastine Fatichah, Bilqis Amaliah, Agus Budi Raharjo, Frédéric Dumont, Emilie Thibaudeau, Cédric Dumas

    A lesion is an abnormal change in body tissue or organ caused by various factors such as inflammation, infection, or abnormal cell growth. The presence of lesions often indicates a serious condition requiring medical attention. Early detection of lesions is crucial for effective treatment and preventing disease progression. Endoscopy is a key tool for lesion identification and evaluation, and the integration of deep learning techniques has further enhanced its potential. This systematic literature review (SLR) examines 35 studies to provide a comprehensive overview of lesion detection using deep learning on endoscopic images. The review explores deep learning techniques, modifications to baseline models, datasets, preprocessing and augmentation methods, and evaluation metrics. A meta-analysis of average performance metrics—accuracy, precision, recall, and specificity—was also conducted to identify trends across datasets and lesion types. Additionally, emerging trends such as federated learning, lightweight models, and ethical considerations are briefly discussed, addressing critical aspects for clinical translation. While significant advancements have been made, challenges remain, including dataset biases, scalability in resource-limited settings, and ensuring generalizability. Insights gained from this review are expected to guide future developments in lesion detection models, contributing to improved diagnostic accuracy, scalability, and patient outcomes.


  • Comparative Analysis of YOLO-Based Object Detection Models for Peritoneal Carcinomatosis

    by Naim Rochmawati, Chastine Fatichah, Bilqis Amaliah, Agus Budi Raharjo, Frédéric Dumont, Emilie Thibaudeau, Cédric Dumas