High True Positive Rate Research Articles

Bayesian analysis with inaccurate rapid antigen testing for detecting someone is positive

This paper examines how testing accuracy plays a key role in detecting someone is positive. Bayesian-theorem exploits how an inaccurate rapid antigen test with a high true-positive rate and a low false-positive rate affects the detection of the probability that someone is positive. The probability is calculated with three determinants in actual dental practices: Infection rate, true positive rate, and false positive rate. The result suggests that in a high infection rate, the false positive rate of rapid antigen tests plays a key role in detecting positive individuals. In 6.5% infection rate, 0.991 true positive rate, and 0.0005 false positive rate, the probability is 0.992 while it with 0.05 false positive rate is reduced to 0.579. The proposed Bayesian analysis can be used for future analysis with imprecise tests in other applications.

A novel hybrid feature extraction and ensemble C3D classification for anomaly detection in surveillance videos

Anomaly detection in several deep learning frameworks are recently presented on real-time video databases as a challenging task. However, these frameworks have high false positive rate (FPR) and error rate due to various backgrounds, motion appearance and semantic high-level and low-level features for anomaly detection through action classification. Also, extraction of features and classification are the major problems in traditional convolution neural network (CNN) on real-time video databases. The proposed work is a novel action classification framework which is designed and implemented on large video databases with high true positive rate (TPR) and error rate. In this framework, Kalman based incremental principal component analysis (IPCA) feature extraction method; C3D and non-linear support vector machine (SVM) classifier are used to improve the action prediction (anomaly detection) on the large real-time video databases. The proposed frame work shown new results of high computation performance than the traditional deep learning frameworks for action classification.

DenseLens – Using DenseNet ensembles and information criteria for finding and rank-ordering strong gravitational lenses

ABSTRACT Convolutional neural networks (CNNs) are the state-of-the-art technique for identifying strong gravitational lenses. Although they are highly successful in recovering genuine lens systems with a high true-positive rate, the unbalanced nature of the data set (lens systems are rare), still leads to a high false positive rate. For these techniques to be successful in upcoming surveys (e.g. with Euclid) most emphasis should be set on reducing false positives, rather than on reducing false negatives. In this paper, we introduce densely connected neural networks (DenseNets) as the CNN architecture in a new pipeline-ensemble model containing an ensemble of classification CNNs and regression CNNs to classify and rank-order lenses, respectively. We show that DenseNets achieve comparable true positive rates but considerably lower false positive rates (when compared to residual networks; ResNets). Thus, we recommend DenseNets for future missions involving large data sets, such as Euclid, where low false positive rates play a key role in the automated follow-up and analysis of large numbers of strong gravitational lens candidates when human vetting is no longer feasible.

Enhancing Gastrointestinal Disease Diagnosis with KNN: A Study on WCE Image Classification

This study explores the application of the K-Nearest Neighbors (KNN) algorithm, following Sobel segmentation and Hu Moment feature extraction, to classify Wireless Capsule Endoscopy (WCE) images into Normal and Ulcerative Colitis conditions. Through a rigorous 5-fold cross-validation approach, the research aimed to determine the KNN algorithm's accuracy, precision, recall, and F1-score on the WCE Curated Colon Disease Dataset. The findings revealed high performance across all metrics, with accuracy rates extending up to 90.625%. The confusion matrix provided further validation, illustrating a high true positive rate coupled with a low false negative rate. These results substantiate the hypothesis that employing edge detection and shape descriptors as pre-processing techniques can significantly enhance the efficacy of machine learning algorithms in medical image classification. The study’s contribution is twofold: it reaffirms the potential of machine learning in the advancement of medical diagnostics and provides a methodological framework for automated image classification that can assist clinicians. It is recommended that future research extends to broader datasets and explores various algorithms to enhance diagnostic precision. In practice, integrating this research into a clinical decision support system could revolutionize diagnostic processes, offering a non-invasive, accurate, and efficient tool for gastroenterological diagnostics.

MDAP: Module Dependency based Anomaly Prediction

In large-scale distributed systems with multiple interconnected modules, failure of even a single module might have a cascading effect and might result in overall system failure. Naturally, timely identification and resolution of these system-wide anomaly is highly crucial for ensuring robust functioning of a system. However, while many frameworks for detecting anomalous behaviour exist today, very few frameworks are proposed which can leverage those anomalies for ultimately detecting system failures (note that some anomalies might not lead to a failure too). To fill this gap, in our work, we propose MDAP, a system anomaly-prediction framework for large-scale distributed systems. Specifically MDAP made the following contributions: (a) It detects anomalous behaviour in a system by considering behavioural changes of the interdependencies across different modules of the system with only information from system logs. (b) Provides the ability to raise early failure alerts. (c) Finally evaluating with three different datasets, MDAP achieves 39%–52% better accuracy compared to the baseline algorithms in identifying periods when the system is performing abnormally—we achieve a high true positive rate of 70.78% and low false positive rate of 3.56% using information about only a few bugs.

PO-03-056 QUANTIFYING THE FREQUENCY AND ACCURACY OF DEVICE-INDICATED ATRIAL FIBRILLATION EPISODES ACROSS DEVICE MANUFACTURERS

Despite standardization efforts, cardiac implantable electronic devices (CIED) report atrial fibrillation (AF) episodes differently across device manufacturers. Similarly, traditional cardiac rhythm management (CRM) protocols have failed to standardize physician-indicated AF. Thus, few studies have characterized the frequency and accuracy of device-indicated AF. This study quantified and compared physician- (‘true’ AF incidence) and device-indicated AF across two patient segments and by CIED manufacturer. Baptist Health Network in Kentucky uses a CRM care platform (Murj, Santa Cruz, CA) to manage 12,000 patients across 14 clinics. Both device- and physician- indicated AF transmissions in 10,986 CIED patients (age 76+12 yrs) including pacemakers, ICD, CRT-P, CRT-D implants between 10/1/2020-5/31/2022 were assessed. AF transmissions were also analyzed in 3,290 (30% of 10,986 pts) pacer dependent or heart block (HB) patients (defined as AV block or had a non-CRT device with average RV pacing >90%) (age 80+11 yrs). True physician-indicated AF episodes represented 31% & 33% of all pts & HB transmissions corresponding with 72% & 68% all pacing/HB pts. Device-indicated AF represented 32% & 33% all pacing/HB transmissions. Overall, 19% & 20% all pacing/HB transmissions were true positives (TP=device- & physician-indicated AF). Device percentages by manufacturer were 35% Boston Scientific, 29% St. Jude, 27% Medtronic, and 8% Biotronik. Medtronic had the lowest device-indicated AF transmissions/pt, the highest true positive rate at 75% (TPR=true positives / total device-indicated AF transmissions) & highest false negative rate at 30% (FNR=false negatives/ total device-indicated normal transmissions). Biotronik had the highest false positive rate at 50% (FPR= false positives/ total device-indicated AF transmissions) & lowest false negative rate (FNR) at 14%. The mean TPR was 61%+11% per manufacturer. The Murj CRM system allows greater standardization of device- and physician-indicated AF episodes, enabling quantification of device-indication frequency and accuracy. In this study, physician-indicated AF was common (72% all pacing patients; 68% HB patients) with often inaccurate AF device classifications. Alternating high FPR and FNR by manufacturers highlight the optimization challenge of reducing false alarms while maintaining a high TPR. Studies like this provide critical context for clinicians interpreting device-indicated events across manufacturers required by CRM.

Intelligent Edge Based Efficient Disease Diagnosis using Optimization Based Deep Maxout Network

The healthcare model is considered an imperative part of remote sensing of health. Finding the disease requires constant monitoring of patients’ health and the detection of diseases. In order to diagnose the disease utilizing an edge computing platform, this study develops a method called grey wolf invasive weed optimization-deep maxout network (GWIWO-DMN). The proposed GWIWO, which is developed by integrating invasive weed optimization (IWO) and grey wolf optimization (GWO), is used here to train the DMN. The distributed edge computing platform consists of four units, namely monitoring devices, first layer edge server, second layer edge server, and cloud server. The monitoring devices are used for accumulating patient information. The preprocessing and feature selection are performed in the first layer edge server. Here, the preprocessing is carried out using the exponential kernel function. The selection of features is done using Jaro–Winkler distance in the first layer edge server. Then, at the second layer edge server, clustering and classification are carried out using deep fuzzy clustering and DMN, respectively. The proposed GWIWO algorithm is used to do the DMN training. Finally, the cloud server processes the decision fusion. The proposed GWIWO-DMN outperformed with the highest true positive rate (TPR) of 89.2%, highest true negative rate (TNR) of 93.7%, and highest accuracy of 90.9%.

Industrial Experience of Finding Cryptographic Vulnerabilities in Large-scale Codebases

Enterprise environment often screens large-scale (millions of lines of code) codebases with static analysis tools to find bugs and vulnerabilities. Parfait is a static code analysis tool used in Oracle to find security vulnerabilities in industrial codebases. Recently, many studies show that there are complicated cryptographic vulnerabilities caused by misusing cryptographic APIs in Java TM 1 In this paper, we describe how we realize a precise and scalable detection of these complicated cryptographic vulnerabilities based on Parfait framework. The key challenge in the detection of cryptographic vulnerabilities is the high false alarm rate caused by pseudo-influences. Pseudo-influences happen if security-irrelevant constants are used in constructing security-critical values. Static analysis is usually unable to distinguish them from hard-coded constants that expose sensitive information. We tackle this problem by specializing the backward dataflow analysis used in Parfait with refinement insights, an idea from the tool CryptoGuard [ 20 ]. We evaluate our analyzer on a comprehensive Java cryptographic vulnerability benchmark and eleven large real-world applications. The results show that the Parfait-based cryptographic vulnerability detector can find real-world cryptographic vulnerabilities in large-scale codebases with high true-positive rates and low runtime cost.

MANDA: On Adversarial Example Detection for Network Intrusion Detection System

With the rapid advancement in machine learning (ML), ML-based Intrusion Detection Systems (IDSs) are widely deployed to protect networks from various attacks. One of the biggest challenges is that ML-based IDSs suffer from adversarial example (AE) attacks. By applying small perturbations (e.g., slightly increasing packet inter-arrival time) to the intrusion traffic, an AE attack can flip the prediction of a well-trained IDS. We address this challenge by proposing <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MANDA</monospace> , a MANifold and Decision boundary-based AE detection system. Through analyzing AE attacks, we notice that 1) an AE tends to be close to its original manifold (i.e., the cluster of samples in its original class) regardless of which class it is misclassified into; and 2) AEs tend to be close to the decision boundary to minimize the perturbation scale. Based on the two observations, we design <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MANDA</monospace> for accurate AE detection by exploiting inconsistency between manifold evaluation and IDS model inference and evaluating model uncertainty on small perturbations. We evaluate <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MANDA</monospace> on both binary IDS and multi-class IDS on two datasets (NSL-KDD and CICIDS) under three state-of-the-art AE attacks. Our experimental results show that <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">MANDA</monospace> achieves high true-positive rate (98.41%) with a 5% false-positive rate.

Detecting network-based internet censorship via latent feature representation learning

Internet censorship is a phenomenon of societal importance and attracts investigation from multiple disciplines. Several research groups, such as Censored Planet, have deployed large scale Internet measurement platforms to collect network reachability data. However, existing studies generally rely on manually designed rules (i.e., using censorship fingerprints) to detect network-based Internet censorship from the data. While this rule-based approach yields a high true positive detection rate, it suffers from several challenges: it requires human expertise, is laborious, and cannot detect any censorship not captured by the rules. Seeking to overcome these challenges, we design and evaluate a classification model based on latent feature representation learning and an image-based classification model to detect network-based Internet censorship. To infer latent feature representations from network reachability data, we propose a sequence-to-sequence autoencoder to capture the structure and the order of data elements in the data. To estimate the probability of censorship events from the inferred latent features, we rely on a densely connected multi-layer neural network model. Our image-based classification model encodes a network reachability data record as a gray-scale image and classifies the image as censored or not using a dense convolutional neural network. We compare and evaluate both approaches using data sets from Censored Planet via a hold-out evaluation. Both classification models are capable of detecting network-based Internet censorship as we were able to identify instances of censorship not detected by the known fingerprints. Latent feature representations likely encode more nuances in the data since the latent feature learning approach discovers a greater quantity, and a more diverse set, of new censorship instances.

Abstract WP16: Automated Vessel Identification In The Circle Of Willis Using Deep Learning Models To Assist AI Based Lesion Detection In Cerebral Vasculature

Introduction: Vascular diseases are among the top causes of death in the world. Automated approaches which allow comparison of vasculature enable both large population studies of vasculature and assist objective diagnosis of vascular disease. Accurate and robust methods for vessel labeling of angiography are an important step to automatically localize lesions or other features of interest. Hypothesis: using a deep learning approach, a model can be developed to identify cerebral arteries. Methods: We used 152 cerebral TOF-MRA angiograms from three publicly available datasets and automatically segmented vessels using a pre-trained nnU-net model. With the segmented vessels, we manually created reference labels for different cerebral arteries using 3DSlicer. After extracting vessel centerlines from the segmentations using VesselVio, we then associated each centerline point with the reference labels to assign individual arterial segments. Finally, in an ablation-type study, several PointNet++ point cloud models were trained and evaluated, using the centerline coordinates, local vessel radius and/or centerline point connectivity information as inputs. Results: The model trained only on centerline coordinates resulted in a high true positive rate (TPR) of 0.92, while models trained with additional features resulted in significantly higher (p&lt;0.05) TPR of 0.95. Different PointNet++ models favored improved labeling of certain arteries, while the internal carotid artery, and posterior communicating artery were consistently the most and least accurately labeled arteries, respectively. Conclusion: We developed a complete pipeline utilizing a deep learning PointNet++ for automatic cerebral vessel identification that labels vessels based on automatically extracted centerlines. Results show that utilizing vessel radius as an additional feature significantly improved cerebral vasculature labeling. The overall TPR of 0.95 shows that our model can be used for fast and reliable artery labeling of cerebral MRA images.

PhageTailFinder: A tool for phage tail module detection and annotation.

Decades of overconsumption of antimicrobials in the treatment and prevention of bacterial infections have resulted in the increasing emergence of drug-resistant bacteria, which poses a significant challenge to public health, driving the urgent need to find alternatives to conventional antibiotics. Bacteriophages are viruses infecting specific bacterial hosts, often destroying the infected bacterial hosts. Phages attach to and enter their potential hosts using their tail proteins, with the composition of the tail determining the range of potentially infected bacteria. To aid the exploitation of bacteriophages for therapeutic purposes, we developed the PhageTailFinder algorithm to predict tail-related proteins and identify the putative tail module in previously uncharacterized phages. The PhageTailFinder relies on a two-state hidden Markov model (HMM) to predict the probability of a given protein being tail-related. The process takes into account the natural modularity of phage tail-related proteins, rather than simply considering amino acid properties or secondary structures for each protein in isolation. The PhageTailFinder exhibited robust predictive power for phage tail proteins in novel phages due to this sequence-independent operation. The performance of the prediction model was evaluated in 13 extensively studied phages and a sample of 992 complete phages from the NCBI database. The algorithm achieved a high true-positive prediction rate (>80%) in over half (571) of the studied phages, and the ROC value was 0.877 using general models and 0.968 using corresponding morphologic models. It is notable that the median ROC value of 992 complete phages is more than 0.75 even for novel phages, indicating the high accuracy and specificity of the PhageTailFinder. When applied to a dataset containing 189,680 viral genomes derived from 11,810 bulk metagenomic human stool samples, the ROC value was 0.895. In addition, tail protein clusters could be identified for further studies by density-based spatial clustering of applications with the noise algorithm (DBSCAN). The developed PhageTailFinder tool can be accessed either as a web server (http://www.microbiome-bigdata.com/PHISDetector/index/tools/PhageTailFinder) or as a stand-alone program on a standard desktop computer (https://github.com/HIT-ImmunologyLab/PhageTailFinder).

Evaluation of Precision Livestock Technology and Human Scoring of Nursery Pigs in a Controlled Immune Challenge Experiment.

The objectives were to determine the sensitivity, specificity, and cutoff values of a visual-based precision livestock technology (NUtrack), and determine the sensitivity and specificity of sickness score data collected with the live observation by trained human observers. At weaning, pigs (n = 192; gilts and barrows) were randomly assigned to one of twelve pens (16/pen) and treatments were randomly assigned to pens. Sham-pen pigs all received subcutaneous saline (3 mL). For LPS-pen pigs, all pigs received subcutaneous lipopolysaccharide (LPS; 300 μg/kg BW; E. coli O111:B4; in 3 mL of saline). For the last treatment, eight pigs were randomly assigned to receive LPS, and the other eight were sham (same methods as above; half-and-half pens). Human data from the day of the challenge presented high true positive and low false positive rates (88.5% sensitivity; 85.4% specificity; 0.871 Area Under Curve, AUC), however, these values declined when half-and-half pigs were scored (75% sensitivity; 65.5% specificity; 0.703 AUC). Precision technology measures had excellent AUC, sensitivity, and specificity for the first 72 h after treatment and AUC values were &gt;0.970, regardless of pen treatment. These results indicate that precision technology has a greater potential for identifying pigs during a natural infectious disease event than trained professionals using timepoint sampling.

MSCA: An Unsupervised Anomaly Detection System for Network Security in Backbone Network

Anomaly detection is a crucial topic in network security which refers to automatically mining known and unknown attacks or threats. Many detectors have been proposed in the last decade. Nonetheless, a practical solution, which is able to provide a high True Positive Rate (TPR) with an acceptable False Positive Rate (FPR) without any prior information, is still challenging due to the complexity and variability of anomaly pattern. In this article, we propose a novel unsupervised detection system called MSCA which applies multiple sketches, K-means++ unsupervised clustering, and association rule mining to detect traffic anomalies and analyze anomalous features and correlations. It can blindly identify known and unknown traffic anomalies without any labeled traffic or prior signatures about data distribution. Rich traffic data is first aggregated and compacted to traffic flows by sketches, and further detected by the combination of clustering algorithm and voting strategy. Then association rule mining is finally utilized to find the anomalous frequent item-sets and association rules. Numerical experiments on MAWILAB datasets demonstrate that the proposed detection method outperforms other reference unsupervised detection methods. It achieves an accuracy of 99.86%, 99.97%, 97.08%, and 95.19% in overall four detection types including IP and port of source and destination.

Gene selection by incorporating genetic networks into case-control association studies.

Large-scale genome-wide association studies (GWAS) have been successfully applied to a wide range of genetic variants underlying complex diseases. The network-based regression approach has been developed to incorporate a biological genetic network and to overcome the challenges caused by the computational efficiency for analyzing high-dimensional genomic data. In this paper, we propose a gene selection approach by incorporating genetic networks into case-control association studies for DNA sequence data or DNA methylation data. Instead of using traditional dimension reduction techniques such as principal component analyses and supervised principal component analyses, we use a linear combination of genotypes at SNPs or methylation values at CpG sites in a gene to capture gene-level signals. We employ three linear combination approaches: optimally weighted sum (OWS), beta-based weighted sum (BWS), and LD-adjusted polygenic risk score (LD-PRS). OWS and LD-PRS are supervised approaches that depend on the effect of each SNP or CpG site on the case-control status, while BWS can be extracted without using the case-control status. After using one of the linear combinations of genotypes or methylation values in each gene to capture gene-level signals, we regularize them to perform gene selection based on the biological network. Simulation studies show that the proposed approaches have higher true positive rates than using traditional dimension reduction techniques. We also apply our approaches to DNA methylation data and UK Biobank DNA sequence data for analyzing rheumatoid arthritis. The results show that the proposed methods can select potentially rheumatoid arthritis related genes that are missed by existing methods.

Performance of quantum kernel on initial learning process

For many manufacturing companies, the production line is very important. In recent years, the number of small-quantity, high-mix products have been increasing, and the identification of good and defective products must be carried out efficiently. At that time, machine learning is a very important issue on shipping inspection using small amounts of data. Quantum machine learning is one of most exciting prospective applications of quantum technologies. SVM using kernel estimation is one of most popular methods for classifiers. Our purpose is to search quantum advantage on classifier to enable us to classifier in inspection test for small size datasets. In this study, we made clear the difference between classical and quantum kernel learning in initial state and propose analysis of learning process by plotting ROC space. To meet the purpose, we investigated the effect of each feature map compared to classical one, using evaluation index. The simulation results show that the learning model construction process between quantum and classical kernel learning is different in initial state. Moreover, the result indicates that the learning model of quantum kernel is the method to decrease the false positive rate (FPR) from high FPR, keeping high true positive rates on several datasets. We demonstrate that learning process on quantum kernel is different from classical one in initial state and plotting to ROC space graph is effective when we analyse the learning model process.

Hybrid cluster based classification model for high dimensional disease prediction databases

As biomedical databases continue to expand, it becomes increasingly difficult to identify a crucial feature for a classification task due to big data size and sparsity issues. Traditional feature subset models rely on fixed-sized dimensions for the feature ranking and classification process, which is not suitable for addressing concerns with sparsity, missing values, and imbalance in the selection of crucial features for the data classification process. To enhance disease prediction effectiveness, this article proposes a hybrid ensemble feature selection method that employs an advanced cluster-based classification model. The model uses an ensemble of rated features to classify the disease with high accuracy and true positive rate. To improve the effectiveness of tree pruning and classification, we introduce a novel cluster-based classification model. We simulated experimental results using various training datasets to predict accuracy. Our proposed results demonstrate that the gene-chemical disease clustering-based classification framework outperforms traditional methods, statistical metrics, and classification models in terms of optimization.

Multiscale contrast enhancement method for small infrared target detection

Infrared search and track (IRST) systems have been applied extensively in the civil and military fields. However, fast, effective, and robust dim and small infrared target detection remains a challenging research topic. Dim and small targets are easily obscured in different backgrounds or noises, which results in a high false alarm rate. A novel strategy is proposed for small infrared target detection to achieve a high true positive rate and robust detection. First, the prominent edge structures of the original image are obtained by constructing a structure tensor. Subsequently, a Gaussian bandpass filter with multiscale windows is used to obtain an infrared image that contains complex edges and small targets. An infrared image with an enhanced signal-to-clutter ratio is achieved by combining with the two methods, whereby the noise can be effectively reduced. Thereafter, candidate regions that contain the target and a smaller number of edges with higher brightness than that of the target are obtained using relaxed threshold segmentation. Finally, a Gaussian gradient contrast method is proposed in which the highlighted edges are removed so that only small infrared targets remain. Extensive experimental results demonstrate that the proposed method can achieve a significant optimization of the true positive and false alarm rates, as well as improved efficiency and robustness.

A novel approach to detect electricity theft based on conv-attentional Transformer Neural Network

With the establishment of advanced metering infrastructure (AMI), more and more deep learning (DL) methods are being used for electricity theft detection. However, there are still problems such as the construction of a DL model that better fits the theft detection task, sample imbalance in the data set and overfitting of the model which limit the potential of DL models. Therefore, a novel end-to-end solution based on DL to solve these problems for electricity theft detection is proposed in this paper. First, we construct a model based on Transformer Neural Network (TNN), which extracts global features of consumption data by calculating the self-attention between each segment obtained by dividing the whole load sequence, and calculates the relative relationship between features for customer classification. Then, we refine the model to further improve its performance. Conv-attentional module is used to embed the input data and capture the local features in each segment. And we minimize the effect of sample imbalance by choosing a suitable loss function, and address the problem of model overfitting by adding normalization layers, dropout regularization and L2 regularization. In addition, grid search is used to determine the optimal values of the model hyper-parameters. Finally, the performance of the proposed model is verified by experiments using the Irish data set. The results show that our method is able to extract features more efficiently and thus has a higher true positive rate (TPR) with a lower false positive rate (FPR) than other state-of-the-art detectors, and it has strong robustness.

A NOVEL FRAMEWORK TO ESTIMATE MULTIDIMENSIONAL MINIMUM EFFECTIVE DOSES USING ASYMMETRIC POSTERIOR GAIN AND -TAPERING.

In this article we address the problem of estimating minimum effective doses in dose-finding clinical trials of multidimensional treatment. We are motivated by a behavioral intervention trial where we introduce sedentary breaks to subjects with a goal to reduce their glucose level monitored over 8 hours. Each sedentary break regimen is defined by two elements: break frequency and break duration. The trial aims to identify minimum combinations of frequency and duration that shift mean glucose, that is, the minimum effective dose (MED) combinations. The means of glucose reduction associated with the dose combinations are only partially ordered. To circumvent constrained estimation due to partial ordering, we propose estimating the MED by maximizing a weighted product of combinationwise posterior gains. The estimation adopts an asymmetric gain function, indexed by a decision parameter , which defines the relative gains of a true negative decision and a true positive decision. We also introduce an adaptive -tapering algorithm to be used in conjunction with the estimation method. Simulation studies show that using asymmetric gain with a carefully chosen is critical to keeping false discoveries low, while -tapering adds to the probability of identifying truly effective doses (i.e., true positives). Under an ensemble of scenarios for the sedentary break study, -tapering yields consistently high true positive rates across scenarios and achieves about 90% true positive rate, compared to 68% by a nonadaptive design with comparable false discovery rate.