Conventional Search Method Research Articles

HAIGEN: Towards Human-AI Collaboration for Facilitating Creativity and Style Generation in Fashion Design

The process of fashion design usually involves sketching, refining, and coloring, with designers drawing inspiration from various images to fuel their creative endeavors. However, conventional image search methods often yield irrelevant results, impeding the design process. Moreover, creating and coloring sketches can be time-consuming and demanding, acting as a bottleneck in the design workflow. In this work, we introduce HAIGEN (Human-AI Collaboration for GENeration), an efficient fashion design system for Human-AI collaboration developed to aid designers. Specifically, HAIGEN consists of four modules. T2IM, located in the cloud, generates reference inspiration images directly from text prompts. With three other modules situated locally, the I2SM batch generates the image material library into a certain designer-style sketch material library. The SRM recommends similar sketches in the generated library to designers for further refinement, and the STM colors the refined sketch according to the styles of inspiration images. Through our system, any designer can perform local personalized fine-tuning and leverage the powerful generation capabilities of large models in the cloud, streamlining the entire design development process. Given that our approach integrates both cloud and local model deployment schemes, it effectively safeguards design privacy by avoiding the need to upload personalized data from local designers. We validated the effectiveness of each module through extensive qualitative and quantitative experiments. User surveys also confirmed that HAIGEN offers significant advantages in design efficiency, positioning it as a new generation of aid-tool for designers.

Denoiseit: denoising gene expression data using rank based isolation trees

BackgroundSelecting informative genes or eliminating uninformative ones before any downstream gene expression analysis is a standard task with great impact on the results. A carefully curated gene set significantly enhances the likelihood of identifying meaningful biomarkers.MethodIn contrast to the conventional forward gene search methods that focus on selecting highly informative genes, we propose a backward search method, DenoiseIt, that aims to remove potential outlier genes yielding a robust gene set with reduced noise. The gene set constructed by DenoiseIt is expected to capture biologically significant genes while pruning irrelevant ones to the greatest extent possible. Therefore, it also enhances the quality of downstream comparative gene expression analysis. DenoiseIt utilizes non-negative matrix factorization in conjunction with isolation forests to identify outlier rank features and remove their associated genes.ResultsDenoiseIt was applied to both bulk and single-cell RNA-seq data collected from TCGA and a COVID-19 cohort to show that it proficiently identified and removed genes exhibiting expression anomalies confined to specific samples rather than a known group. DenoiseIt also showed to reduce the level of technical noise while preserving a higher proportion of biologically relevant genes compared to existing methods. The DenoiseIt Software is publicly available on GitHub at https://github.com/cobi-git/DenoiseIt

High-Speed Power Allocation in NOMA System Using FPGA-Based DNN

Artificial Intelligence (AI) is rapidly transforming the healthcare, finance and transportation industries. This paper presents a field-programmable gate array (FPGA)-based neural network accelerator (NNA) design for power allocation in downlink nonorthogonal multiple access (NOMA) networks. The proposed hardware accelerator effectively cuts computational costs while delivering performance on par with the highest sum capacity. Numerical results show that this NNA offers a remarkable computational speed increase of up to 99% compared to the conventional exhaustive search method. Furthermore, the deep learning (DL) model achieved high accuracy (0.92 training, 0.93 testing), and the hardware accelerator design for this DL inference model was implemented on the PYNQ-Z2 board-constrained edge device to predict power allocation coefficients in NOMA systems.

Hierarchical blind phase search for correcting imperfect phase rotation in QAM signals synthesized via optical coherent superposition.

Coherent superposition has been proposed to synthesize high-order quadrature amplitude modulation (QAM) by coherently superposing low-order QAMs in the optical domain. These approaches could effectively relax the digital-to-analog converter resolution and reduce the complexity of the driving electronics. However, in the superposition process, imperfect phase rotations (IPRs) in low-order QAMs will be transferred to the resultant high-order QAM. Importantly, the induced IPR cannot be compensated for by conventional linear equalizers and carrier recovery methods. To combat the induced IPR, herein, we propose a hierarchical blind phase search (HBPS) algorithm to compensate for the IPRs in synthesized high-order QAMs. The proposed HBPS can match the generation mechanism of the IPRs in coherent superposition, by tracing back and estimating the IPR in the QPSK-like constellation of each hierarchy and finally correcting the induced IPRs. Simulation and experimental results verify that this algorithm could effectively compensate for the IPR in the resultant 16-QAMs synthesized using coherent superposition approaches. The proposed HBPS shows significant optical signal-to-noise ratio (OSNR) gains compared to the conventional blind phase search (BPS) method for high-order QAMs coherently superposed using optical signal processing (OSP) and tandem modulators (TMs). Specifically, at the BER of 2.4e-2, the HBPS achieves a 1.5-dB OSNR sensitivity enhancement over the BPS in either OSP or TMs-based schemes, even with an imperfection rotation of up to 20∘.

Age of Information for Partial Earliest Relay Aided Short Packet Status Update With Energy Harvesting

This paper focuses on the timely delivery of status updates for a multi-relay assisted energy harvesting (EH) Internet of Things system with short packet communications (SPC), where an EH-powered source transmits status updates to a destination with the aid of multiple EH-powered relays. To improve the energy efficiency and information freshness, an earliest-ℓ partial relay selection scheme is proposed, in which the earliest ℓ relays that successfully receive the status update from the source will be selected to forward the status update to the destination, and the selection combining (SC) and maximal-ratio combining (MRC) schemes are used at the destination. Firstly, we investigate the update packet delivery schemes of source-relay (S-R) and relay-destination (R-D) links, and present the packet error probabilities under the SC and MRC schemes based on SPC theory. Secondly, after characterizing the full charge time durations of both the source and relay, we investigate the statistical description of the number of the packet transmission attempts from the source to relay as well as the one from the relay to destination by using order statistics. With the analysis of the evolution of the instantaneous age of information (AoI), the expression for average AoI is derived. Finally, we investigate the average AoI minimization, which is formulated as a coupled triplet problem of the packet lengths at the source and relays and the value of ℓ. To overcome the coupling, we present the convexity approximation of the average AoI as well as the concise proof. Then, an efficient gradient-based two-step optimal search algorithm is proposed to solve the optimization problem, which searches the local optimal packet lengths for S-R and R-D links by iteratively updating the value of ℓ until the global optimal objective parameters are found. Numerical analyses give the effect of the key parameters on the average AoI, and the comparison demonstrates that the proposed gradient-based method converges quickly than the conventional greedy search method.

A Multi-Stage Real-Time Fast Search Method for Closed-Loop Paths Based on Grid Hierarchical Partitioning Characteristics

Prior to switching operations, the dispatch automation system is required to be able to search for closed-loop paths quickly in real time. In order to improve the efficiency of closed-loop path search, this paper proposes a multi-stage real-time fast search method for closed-loop paths based on the characteristics of power grid hierarchical partitioning. First, we divide the closed-loop judgement process into three phases: initialization, acceptance of the closed-loop judgement command, and post-operation. Then, we define three types of nodes, including root nodes, same-layer contact nodes, and common nodes, and construct path data models for each type of node in accordance with the node liaison relationship in grid hierarchical zoning. In the initialization phase, we design a method for the automatic generation of real-time nodal path data models and study a hierarchical partitioned closed-loop path search method based on a nodal path data model for the phase of accepting closed-loop judgement commands. Next, we design a fast local correction method for nodal path models after the switch closure and disconnection operations in the post-operation idle phase are performed. Finally, a real power grid model is used as an example to test the above closed-loop path search method, which improves the search efficiency by eight times compared with the conventional breadth-first search method. The method is applicable to real large and complex power grids.

The Alzheimer's Disease Neuroimaging Initiative in the era of Alzheimer's disease treatment: A review of ADNI studies from 2021 to 2022.

The Alzheimer's Disease Neuroimaging Initiative (ADNI) aims to improve Alzheimer's disease (AD) clinical trials. Since 2006, ADNI has shared clinical, neuroimaging, and cognitive data, and biofluid samples. We used conventional search methods to identify 1459 publications from 2021 to 2022 using ADNI data/samples and reviewed 291 impactful studies. This review details how ADNI studies improved disease progression understanding and clinical trial efficiency. Advances in subject selection, detection of treatment effects, harmonization, and modeling improved clinical trials and plasma biomarkers like phosphorylated tau showed promise for clinical use. Biomarkers of amyloid beta, tau, neurodegeneration, inflammation, and others were prognostic with individualized prediction algorithms available online. Studies supported the amyloid cascade, emphasized the importance of neuroinflammation, and detailed widespread heterogeneity in disease, linked to genetic and vascular risk, co-pathologies, sex, and resilience. Biological subtypes were consistently observed. Generalizability of ADNI results is limited by lack of cohort diversity, an issue ADNI-4 aims to address by enrolling a diverse cohort.

Search Method of Robust Gate Driving Vectors for Digital Gate Drivers With Low Test Cost Against Load Current and Temperature Variations in IGBTs

A digital gate driver with active gate driving is effective in solving the trade-off between switching loss and current/voltage overshoot of power devices. However, the optimum gate driving vector (GV) for a digital gate driver is temperature- and load current-dependent. When an optimum GV at a particular operating condition is reused to another, the switching performance sometimes becomes worse than the conventional gate drive [17]. Therefore, robust gate driving vectors (RGV) against temperature and load current variation are required [18]. Nevertheless, the test cost of searching for an RGV is very high. In order to reduce the test cost, search methods of RGV with fewer required measurements are proposed in this work. Single-step GV and stop-and-go GV are selected as the type of RGV for turn-on and turn-off, respectively. Compared with the conventional search method of RGV in [18] and [20], the proposed search method can reduce the measurement time by 99 % and 92 %, respectively. Furthermore, when the RGV and the optimum GV at a single condition are reused to other operating conditions, the maximum overall <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">f</i> <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">OBJ</sub> of RGV is lower than that of the single condition optimum GV, which shows that the searched RGV has better robustness.

Identification of 113 new histone marks by CHiMA, a tailored database search strategy.

Shotgun proteomics has been widely used to identify histone marks. Conventional database search methods rely on the "target-decoy" strategy to calculate the false discovery rate (FDR) and distinguish true peptide-spectrum matches (PSMs) from false ones. This strategy has a caveat of inaccurate FDR caused by the small data size of histone marks. To address this challenge, we developed a tailored database search strategy, named "Comprehensive Histone Mark Analysis (CHiMA)." Instead of target-decoy-based FDR, this method uses "50% matched fragment ions" as the key criterion to identify high-confidence PSMs. CHiMA identified twice as many histone modification sites as the conventional method in benchmark datasets. Reanalysis of our previous proteomics data using CHiMA led to the identification of 113 new histone marks for four types of lysine acylations, almost doubling the number of previously reported marks. This tool not only offers a valuable approach for identifying histone modifications but also greatly expands the repertoire of histone marks.

Task weighting based on particle filter in deep multi-task learning with a view to uncertainty and performance

Recently multi-task learning (MTL) has been widely used in different applications to build more robust models by sharing knowledge across several related tasks. However, one challenge that arises is the variability in the learning pace of different tasks causing the inefficiency of naively training all tasks. Therefore, it is of great importance to consider some coefficients to balance tasks in the process of learning, but, due to the large search space and the significance of setting them properly, conventional search methods such as grid or random search are no longer effective. In this paper, we propose a learning mechanism for these coefficients based on the high efficiency of the particle filter (PF) algorithm to deal with nonlinear search problems. PF considers each state of the tasks’ coefficients as a particle and recursively converges coefficients to an optimum point. While in most previous works coefficients were evaluated to only increase performance, to address the recent concerns related to applying AI in real-world applications, we also incorporate uncertainty alongside our method to prevent learning coefficients leading to unstable outcomes. This mechanism is independent of the models main learning process and can be easily added to every learning system without changing its training algorithm. Extensive experiments on real-world data sets demonstrate the superiority of the proposed method over the state-of-the-art methods on both performance and uncertainty. We also proved the acceptable performance of the method using Cramer Rao lower bound theory.

Closed-Loop Estimation of Individualized Inter-Stimulus Interval Window for Transient Neuromodulation via Paired Mechanical and Brain Stimulation

Mechanical stimulation-conditioned transcranial magnetic stimulation (Mstim-cTMS) is a neuromodulation technique that transiently enhances neural excitability through paired peripheral and cortical stimulation. Importantly, the temporal inter-stimulus interval (ISI) must fall within an individual’s specific ISI window (ISI-W) to induce neuromodulatory effects. Currently, ISI-W identification is inefficient, incrementally applying Mstim-cTMS within a wide search range. This paper aims to test a real-time, closed-loop framework that decreases the number of trials required to obtain individualized ISI-Ws by implementing a previously developed statistical regression algorithm. Each Mstim-cTMS trial consisted of peripheral mechanical stimulation using a precise, tapping robot paired with transcranial magnetic stimulation to the motor cortex. Neuromodulation was characterized by measuring resulting motor evoked potentials (MEPs). The closed-loop process estimated the effective ISI-W by iterating Mstim-cTMS at statistically determined ISIs based on the detected MEPs. The estimated ISI-Ws of two subjects had an average correlation coefficient of 0.84 with their reference windows obtained using the conventional search method. The estimation reduced the number of required Mstim-cTMS trials by an average of 93.8%. The results indicate the proposed method’s potential for decreasing Mstim-cTMS startup durations. Future rehabilitation and neurophysiological studies can benefit from efficiently updating ISI-Ws to account for long-term neural variability.

Pilot trial comparing COVID-19 publication database to conventional online search methods

Background and objectivesLiterature review using search engines results in a list of manuscripts but does not provide the content contained in the manuscripts. Our goal was to evaluate user performance-based...

Design and Implementation of a Medical Question and Answer System Based on Deep Learning

Medical services play a pivotal role in people’s lives and in the national economy. Although the number of healthcare facilities is currently growing every year, there are still major problems in terms of access and pressure on the flow of people. Therefore, there is an urgent need for complementary medical services to alleviate the flow of patients and their psychological burden and to enable them to receive timely medical advice. This article designs and implements a medical Q&amp;A system based on deep learning. We took a retrieval-based approach, using crawler technology that has been manually reviewed to build the Q&amp;A database, and the Seq2Seq algorithm and the TF-IDF model to build the answer generation model. The medical question and answer system developed enable effective Q&amp;A and relevant medical advice to be given. The algorithm proposed in this paper can quickly provide users with accurate answers compared to conventional search methods in real datasets.

NURBS Surface-Altering Optimization for Identifying Critical Slip Surfaces in 3D Slopes

The evaluation of slope stability of a three-dimensional slope requires identifying the critical slip surface with the minimum factor of safety, which is a complex optimization problem. Failure to identify the critical slip surface can lead to unconservative conclusions about the stability of a slope. This paper proposes a novel 3D surface-altering optimization method, which iteratively alters the geometry of a 3D slip surface to find the critical slip surface representing the minimum factor of safety in a slope. The geometry of the slip surface is defined via nonuniform rational basis spline (NURBS) curves formed over a plan grid of control points. The proposed method includes a series of five subroutines that apply various forms of transformations to the control points. These subroutines include minimization problems, which determine the optimal transformation parameters for minimizing the obtained factor of safety of the resulting slip surfaces. Given that any geometrically defined slip surface can be approximated using an equivalent series of NURBS control points, the proposed method can be used in efforts to further reduce the global factor of safety first obtained via conventional search methods, such as those involving spherical or ellipsoidal slip surfaces. To demonstrate its effectiveness, the proposed method was applied to further optimize the critical ellipsoidal slip surfaces reported in some numerical examples. Comparing the results with those limited to ellipsoidal slip surfaces, the proposed method was consistently able to identify slip surfaces with significantly lower factors of safety. The postaltered slip surfaces also matched closely with finite element shear strength reduction results. As such, the proposed method is effective in searching for critical slip surfaces and can be used as a final step in the critical surface searching routine.

Improved TDV algorithm for three-dimensional space path planning in a complex radioactive environment with obstacles

Decommissioning of nuclear power plants is a complex process that exposes workers to radiation risks. Recently, many shortest path algorithms have been proposed to find a minimum dose path and the applications are generally based on three concepts: environmental modeling, radiation dose computation, and shortest path algorithm. However, the conventional minimum dose path search methods are sub-optimal because they suffer from semi-empirical approximation, complex geometries restrictions, and lack integrated environmental modeling capability. This study presents an improved three-degree of vertex (iTDV) algorithm, an extension of the novel three-degree of vertex (TDV) shortest path algorithm previously developed. In this work, we enhance its search mechanism to find the shortest path with the least amount of radiation dose in a complex three-dimensional (3D) space. The work also outlines the effective coupling of TDV with FLUKA Monte Carlo environmental modeling, which prevents interpolation uncertainty associated with the conventional path planning methods when dealing with large geometries. To demonstrate its efficiency, the iTDV is used to find the minimum dose walking path in environments with and without obstacles. The performance of the iTDV algorithm is also compared with the original TDV, and the Dijkstra's algorithm. The evaluation results and the improvements of the proposed method are discussed in this paper.

Research on Search Algorithm Using Particle Swarm Optimization with Virtual Pheromone for Swarm Robots

This paper proposes a search algorithm using particle swarm optimization (PSO) with virtual pheromone for swarm robots. Swarm robots are attracting attention in disaster relief works to search for victims. The search algorithm involves a combination of global and local searching. The conventional search method consists of random walk as the global search and PSO as the local search. However, random walk is not efficient in complex environments. For efficient searching, PSO with virtual pheromone is used for the global search. The virtual pheromone drives the swarm robots to an unsearched area, dose not need map data, and has low calculation cost. In addition, it is not necessary in the proposed method to switch algorithms between global and local searching. The validity of the proposed method was confirmed from the simulation results.

A Novel Network Resolved and Mobile Assisted Cell Search Method for 5G Cellular Communication Systems

When the user equipment (UE) is turned on, it first needs to perform an initial cell search procedure before it can communicate with the base station (BS). However, this procedure must deal with the initial time and frequency synchronization and search for the appropriate home BS for services. In conventional cell search methods, the entire cell search process is carried out on the UE, where the computational requirement can lead to additional overhead and power consumption of the UE. Therefore, aiming to help UEs offloading their overhead to BSs, we in this paper propose a novel cell search method, called <i>Network Resolved and Mobile Assisted Cell Search</i>, which can achieve our goal by letting the BSs be the main performers for deciding the appropriate BSs for connection and letting UEs only be the assistants of the cell search process. The implementation details and analysis of the proposed cell search approach are provided along with the numerical simulations. Results show that our proposed cell search not only can offload the computational overhead, but it also outperforms the conventional code-based cell search approach in terms of the cell search error probability with similar overall network energy consumption.

Using the Alzheimer's Disease Neuroimaging Initiative to improve early detection, diagnosis, and treatment of Alzheimer's disease.

IntroductionThe Alzheimer's Disease Neuroimaging Initiative (ADNI) has accumulated 15 years of clinical, neuroimaging, cognitive, biofluid biomarker and genetic data, and biofluid samples available to researchers, resulting in more than 3500 publications. This review covers studies from 2018 to 2020.MethodsWe identified 1442 publications using ADNI data by conventional search methods and selected impactful studies for inclusion.ResultsDisease progression studies supported pivotal roles for regional amyloid beta (Aβ) and tau deposition, and identified underlying genetic contributions to Alzheimer's disease (AD). Vascular disease, immune response, inflammation, resilience, and sex modulated disease course. Biologically coherent subgroups were identified at all clinical stages. Practical algorithms and methodological changes improved determination of Aβ status. Plasma Aβ, phosphorylated tau181, and neurofilament light were promising noninvasive biomarkers. Prognostic and diagnostic models were externally validated in ADNI but studies are limited by lack of ethnocultural cohort diversity.DiscussionADNI has had a profound impact in improving clinical trials for AD.

Optimizing surfactant-enhanced aquifer remediation based on Gaussian process surrogate model in DNAPL-contaminated sites considering different wells patterns

Surfactant-enhanced aquifer remediation (SEAR) is an appropriate method for Dense non-aqueous phase liquids (DNAPLs) remediation. However, due to the high cost of chemicals used, choosing the suitable wells pattern and the optimal pumping scenario is necessary. In this study, the SEAR method performance for Regular (convergent) and Inverted (divergent) patterns with different wells numbers have been evaluated. The performance of 5 categories of patterns, including 35 different sub-patterns, was evaluated in a PCE-contaminated aquifer. The results show that the uniformity and appropriate surfactant distribution in the contaminated area significantly improves remediation performance. The distribution of surfactants in Regular patterns was better than Inverted patterns, and Regular patterns had lower remediation duration and cost. The best patterns that achieved a 95 % removal rate at the lowest cost were Regular. To find the optimal pumping scenario, a simulation-optimization model based on the Gaussian process regressor (GPR), as a surrogate model, has been used to reduce the optimization model's computational burden. Nine different kernels were applied and evaluated to find the best GPR. Also, the Bayesian hyperparameter optimization (BHO) method was used to optimize the surrogate model, and its performance was compared with the conventional grid search method. The results showed that the use of the Chi2 kernel and the BHO method are the best choices. A BHO-optimized multi-kernel Gaussian process (BHOMK-GP) model has also been developed, and its performance has been compared with single-kernel GPR surrogate models. The BHOMK-GP model's accuracy was significantly higher than single-kernel GPR models. The test and cross-validation RMSE of the BHOMK-GP model were 0.0385 and 0.0435, respectively. Finally, the optimal remediation scenario has been obtained by substituting the BHOMK-GP model as a surrogate model instead of the SEAR simulation model. The cost of remediation in the optimal strategy was $ 77,575.

Development of Management Information System Reservations Using Simple Multi-Attribute Rating Technique (SMART) Based on Android Mobile Applications

In the information age, technology has become one of the things that are growing more rapidly, one of the technologies is the internet. From the use of the internet, it can also be used by sales that have Android mobile-based bread ordering information systems. By only using a cell phone, system users can order bread. This can be a tool in the special community in ordering bread. As with computers, mobile phones can also install a variety of software or software that they want. Besides, android phones can also be used for transactions. Like making a bread order transactions through an android phone. The stages that will be carried out by the author for the application development process include analysis and needs, application design, and application interface design, so that the application is made easy to use. The ordering system designed can make it easier for admin to view order details, check stock, manage customer data by using a database as a storage medium, and can also provide customers to buy stock of bakery goods available at these outlets to provide accurate information. The SMART method has been proven to be implemented in searches based on user preferences, and gives different results if the value of the criteria entered is different. However, conventional search methods are still provided in addition to searching for goods using the SMART Method.