Design Of Algorithms Research Articles

Latency-Aware Unified Dynamic Networks for Efficient Image Recognition.

Dynamic networks have become a pivotal area of study in deep learning due to their ability to selectively activate computing units (such as layers or channels) or dynamically allocate computation to information-rich regions. This capability significantly curtails unnecessary computations, adapting to varying inputs. Despite these advantages, the practical efficiency of dynamic models often falls short of theoretical computation. This discrepancy arises from three primary challenges: 1) a lack of a unified framework across different dynamic inference paradigms due to the fragmented research landscape; 2) an excessive focus on algorithm design at the expense of scheduling strategies, which are essential for optimizing resource utilization on hardware; and 3) the complexity of latency evaluation, since most current libraries cater to static operators. To tackle these issues, we introduce Latency-Aware Unified Dynamic Networks (LAUDNet), a general framework that integrates three fundamental dynamic paradigms-spatially-adaptive computation, layer skipping, and channel skipping-into a single unified formulation. LAUDNet not only refines algorithmic design but also enhances scheduling optimization with the aid of a latency predictor. This predictor efficiently and accurately predicts the inference latency of dynamic operators on specific hardware setups. Our empirical assessments across multiple vision tasks-image classification, object detection, and instance segmentation-confirm that LAUDNet significantly bridges the gap between theoretical and practical efficiency. For instance, LAUDNet cuts down the practical latency of its static counterpart, ResNet-101, by over 50% on hardware platforms like V100, RTX 3090, and TX2 GPUs. Additionally, LAUDNet excels in the accuracy-efficiency trade-off compared to other methods.

BEATRIX: An open source humanoid head platform for robotics teaching and research

This paper introduces BEATRIX, a novel robotic head designed to bridge the gap between theoretical knowledge and practical experience in the field of robotics at universities. The BEATRIX robot comprises a head actuated by a neck-like mechanism with three stepper motors, two cameras and two microphones for acquisition of visual and audio information from the environment. The robot can be connected to any external computer for the design and implementation of algorithms for applications in human–robot interaction. The proposed robotic platform has been used successfully with undergraduate and master students implementing tasks such as face detection and tracking, sound detection and tracking, robot control and graphical user interfaces. This paper includes lists of all the robot components, assembly instructions, and links to all CAD and software files, facilitating replication and further exploration. The robot design and integration of visual and audio sensors enables the development of engaging educational tutorials and robot experiments, enhancing the teaching and learning experience in robotics.

GACT-PPIS: Prediction of protein-protein interaction sites based on graph structure and transformer network

The prediction of protein-protein interaction sites (PPIS) is currently crucial for regulating many biological activities in cells and developing drugs for various diseases. Deep learning-based methods have been proposed for predicting PPIS, significantly reducing the manpower and time costs associated with traditional experimental methods such as yeast two-hybrid, mass spectrometry, and affinity purification. However, the predictive accuracy of these deep learning methods has not yet reached the expected level. Therefore, we introduce a model called GACT-PPIS.The design of the GACT-PPIS algorithm aims to utilize combined information from protein sequences and structures as input to predict protein-protein interaction sites. The core of GACT-PPIS utilizes an Enhanced Graph Attention Network (EGAT) with initial residual and identity mappings, along with a deep Transformer network as the basic units, supplemented by Graph Convolutional Networks (GCN), effectively aggregating information from neighboring nodes for each node. After multiple network layers, the information of the entire protein is also fused into the nodes, and the Transformer network further enhances the model's performance.Experimental results show that GACT-PPIS outperforms the most representative models in terms of Recall, F1-measure, MCC, AUROC, and AUPRC on the benchmark test set (Test-60). Additionally, on other independent test sets (UBTest-31-6), GACT-PPIS leads in terms of Accuracy, Precision, Recall, F1-measure, MCC, AUROC, and AUPRC compared to the most representative models. It is worth noting that GACT-PPIS demonstrates excellent generalization and versatility across different test sets, showcasing good performance on multiple test sets for the same trained GACT-PPIS model.

Distributed Algorithm Design for Nonsmooth and Nonlinear Resource Allocation Problems of Autonomous High-Order Agents and Its Application to Smart Grids

Distributed Algorithm Design for Nonsmooth and Nonlinear Resource Allocation Problems of Autonomous High-Order Agents and Its Application to Smart Grids

Safe Reinforcement Learning With Dual Robustness.

Reinforcement learning (RL) agents are vulnerable to adversarial disturbances, which can deteriorate task performance or break down safety specifications. Existing methods either address safety requirements under the assumption of no adversary (e.g., safe RL) or only focus on robustness against performance adversaries (e.g., robust RL). Learning one policy that is both safe and robust under any adversaries remains a challenging open problem. The difficulty is how to tackle two intertwined aspects in the worst cases: feasibility and optimality. The optimality is only valid inside a feasible region (i.e., robust invariant set), while the identification of maximal feasible region must rely on how to learn the optimal policy. To address this issue, we propose a systematic framework to unify safe RL and robust RL, including the problem formulation, iteration scheme, convergence analysis and practical algorithm design. The unification is built upon constrained two-player zero-sum Markov games, in which the objective for protagonist is twofold. For states inside the maximal robust invariant set, the goal is to pursue rewards under the condition of guaranteed safety; for states outside the maximal robust invariant set, the goal is to reduce the extent of constraint violation. A dual policy iteration scheme is proposed, which simultaneously optimizes a task policy and a safety policy. We prove that the iteration scheme converges to the optimal task policy which maximizes the twofold objective in the worst cases, and the optimal safety policy which stays as far away from the safety boundary. The convergence of safety policy is established by exploiting the monotone contraction property of safety self-consistency operators, and that of task policy depends on the transformation of safety constraints into state-dependent action spaces. By adding two adversarial networks (one is for safety guarantee and the other is for task performance), we propose a practical deep RL algorithm for constrained zero-sum Markov games, called dually robust actor-critic (DRAC). The evaluations with safety-critical benchmarks demonstrate that DRAC achieves high performance and persistent safety under all scenarios (no adversary, safety adversary, performance adversary), outperforming all baselines by a large margin.

Automatic design generation of trusses from a reused steel stock library using graphic statics

Reuse of undamaged steel components in new structural design offers a means to reduce construction-related waste and carbon emissions. However, designing with reused members significantly differs from conventional methods as the design must conform to a so-called stock constraint. This work presents an automated stock-constrained design algorithm for planar trusses. To promote high efficiency of the resulting designs, graphic statics is used as the backbone of the algorithm. Additionally, optimization is used to reduce the wasted stock material when determining how to cut stock members to form the truss. To address the need for flexibility in reuse design, the underlying engine has inherent stochasticity and is fast (on the order of 1 min). It is herein executed 100 times per design case and shown to generate efficient design options that can either construct the same truss layout in multiple ways or provide the designer with different layout alternatives to select from. Carbon savings are estimated as 83%–90% for an obtained design case when compared to conventionally designed trusses made from recycled steel.

Countering flaws in algorithm design and applications: a Delphi study

Countering flaws in algorithm design and applications: a Delphi study

Designing proteins: Mimicking natural protein sequence heterogeneity.

This study presents an enhanced protein design algorithm that aims to emulate natural heterogeneity of protein sequences. Initial analysis revealed that natural proteins exhibit a permutation composition lower than the theoretical maximum, suggesting a selective utilization of the 20-letter amino acid alphabet. By not constraining the amino acid composition of the protein sequence but instead allowing random reshuffling of the composition, the resulting design algorithm generates sequences that maintain lower permutation compositions in equilibrium, aligning closely with natural proteins. Folding free energy computations demonstrated that the designed sequences refold to their native structures with high precision, except for proteins with large disordered regions. In addition, direct coupling analysis showed a strong correlation between predicted and actual protein contacts, with accuracy exceeding 82% for a large number of top pairs (>4L). The algorithm also resolved biases in previous designs, ensuring a more accurate representation of protein interactions. Overall, it not only mimics the natural heterogeneity of proteins but also ensures correct folding, marking a significant advancement in protein design and engineering.

Optically Transparent Complex-Amplitude Metasurface for Full-Space Manipulation of Frequency-Multiplexed Holographic Imaging.

Restricted by the physical properties of materials, most traditional metasurfaces (MSs) cannot achieve transparent stealth in the visible spectrum. Although some metasurfaces for holography have been designed, there is no standard method for evaluating the advancement of wavefront manipulation under different design algorithms. Here, a complex-amplitude metasurface with optical transparency (OT) and full-space manipulation is presented in the millimeterwave band. Through frequency multiplexing, two holographic images "HOLO" and "GRAM" with high quality are designed, achieving the signal-to-noise ratios of 14.6 and 14.4 dB, respectively. The 64 × 64 bilayered metasurface consists of glass and poly(ethylene terephthalate) (PET) substrates and three metallic mesh layers. By change of the opening size and oriented angle of split rings, phase and amplitude modulations are independently realized. Besides, a new method is provided to assess the superiority of the metasurface holography system for near-field energy regulation. The prototype was fabricated using micro-nano-technology and tested. Two holographic images were experimentally demonstrated and show good agreement with theoretical calculations and simulated results. With the advantages of OT and full-space complex-amplitude customization, the proposed metasurface holograms for frequency-multiplexed may find prospective application aspects in target perception, multichannel data storage and encryption, and many other related fields.

ELLIPSOIDAL DESIGN OF SLIDING MODE CONTROL FOR CAR ACTIVE SUSPENSION SYSTEM

The issue of state feedback sliding mode control (SMC) for a particular kind of active quarter-car suspension system is discussed in this work. The suspension systems' dynamic system is built with the three control objectives of maximal actuator control force, suspension deflection, and ride comfort in mind. The next step is to build the state-feedback controller with the disturbance (vibration) attenuation level in mind to guarantee the asymptotic stability of the closed-loop system. The algorithm for SMC design is introduced. It is predicated on choosing the sliding surface correctly using the invariant ellipsoid approach. The system motion in the sliding mode is guaranteed to be little affected by mismatched disturbances according to the control architecture. Furthermore, the presence of admissible controllers is articulated in terms of LMIs through the use of Lyapunov theory and the linear matrix inequality (LMI) technique. The objective is to create a desired dynamic state-feedback controller when these requirements are met. Lastly, a quarter-car model is used to illustrate how successful a suggested approach is.

On Process Discovery Experimentation

Process mining aims to derive insights into business processes from event logs recorded from information systems. Process discovery algorithms construct process models that describe the executed process. With the increasing availability of large-scale event logs, process discovery has shifted towards a data-oriented research discipline, aiming to design algorithms that are applicable and useful in practice. This shift has revealed a fundamental problem in process discovery research: Currently, contributions can only be considered in isolation. Researchers conduct experiments to show that they move the field forward, but due to a lack of reliability and validity, the individual contributions are hard to generalise. In this paper, we argue that one reason for these problems is the lack of conventions or standards for experimental design in process discovery. Hence, we propose “process discovery engineering”: a research methodology for process discovery, consisting of a shared terminology and a checklist for conducting experiments. We demonstrate its applicability by means of an example experimental evaluation of process discovery algorithms and discuss the implications of the methodology on the field. This paper is not meant to be prescriptive, but to invite and encourage the community to contribute to this discussion to advance the field as a whole.

Average Case Subquadratic Exact and Heuristic Procedures for the Traveling Salesman 2-OPT Neighborhood

We describe an exact algorithm for finding the best 2-OPT move that, experimentally, was observed to be much faster than the standard quadratic approach for a large part of a best-improvement local search convergence starting at a random tour. To analyze its average-case complexity, we introduce a family of heuristic procedures and discuss their complexity when applied to a random tour in graphs whose edge costs are either uniform random numbers in [0, 1] or Euclidean distances between random points in the plane. We prove that, for any probability p, there is a heuristic in the family that can find the best 2-OPT move with probability at least p in average-time [Formula: see text]) for uniform instances and [Formula: see text] for Euclidean instances. The exact algorithm is then proved to be even faster in the sense that in those instances in which a heuristic finds the best move, the exact algorithm finds it in a smaller time. We give empirical evidence that a slight variant of our algorithm finds the best move in [Formula: see text] time on both types of instances, achieving the best possible performance for this particular problem. Computational experiments are reported to show the effectiveness of our algorithms, both in best-improvement and in first-improvement 2-OPT local search. History: Accepted by Andrea Lodi, Area Editor for Design and Analysis of Algorithms: Discrete. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.0169 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2023.0169 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

The Design of a Script Identification Algorithm and Its Application in Constructing a Text Language Identification Dataset

Script identification is easier to implement than language identification, and its identification rate is very high. The fewer languages are identified when using a language identification algorithm, the higher the identification rate is. However, no systematic study on SI involving multiple languages and determining how to construct relevant language identification datasets has been conducted. Therefore, in this paper, we discuss and design a script identification algorithm and the construction of a language identification dataset based on script groups. The data sources in this paper comprise 261 different languages’ text corpora from the Leipzig Corpora Collection, which are grouped into 23 different script groups. In the Unicode encoding scheme, different scripts are arranged into different code regions. Based on this feature, we propose a written script identification algorithm based on regular expression matching, the micro F-score of which reaches 0.9929 in sentence-level script identification experiments. To reduce noise when constructing the language identification dataset for each script, a script identification algorithm is used to filter out other-script content in each text.

A computational algorithm for optimal design of a bioartificial organ scaffold architecture.

We develop a computational algorithm based on a diffuse interface approach to study the design of bioartificial organ scaffold architectures. These scaffolds, composed of poroelastic hydrogels housing transplanted cells, are linked to the patient's blood circulation via an anastomosis graft. Before entering the scaffold, the blood flow passes through a filter, and the resulting filtered blood plasma transports oxygen and nutrients to sustain the viability of transplanted cells over the long term. A key issue in maintaining cell viability is the design of ultrafiltrate channels within the hydrogel scaffold to facilitate advection-enhanced oxygen supply ensuring oxygen levels remain above a critical threshold to prevent hypoxia. In this manuscript, we develop a computational algorithm to analyze the plasma flow and oxygen concentration within hydrogels featuring various channel geometries. Our objective is to identify the optimal hydrogel channel architecture that sustains oxygen concentration throughout the scaffold above the critical hypoxic threshold. The computational algorithm we introduce here employs a diffuse interface approach to solve a multi-physics problem. The corresponding model couples the time-dependent Stokes equations, governing blood plasma flow through the channel network, with the time-dependent Biot equations, characterizing Darcy velocity, pressure, and displacement within the poroelastic hydrogel containing the transplanted cells. Subsequently, the calculated plasma velocity is utilized to determine oxygen concentration within the scaffold using a diffuse interface advection-reaction-diffusion model. Our investigation yields a scaffold architecture featuring a hexagonal network geometry that meets the desired oxygen concentration criteria. Unlike classical sharp interface approaches, the diffuse interface approach we employ is particularly adept at addressing problems with intricate interface geometries, such as those encountered in bioartificial organ scaffold design. This study is significant because recent developments in hydrogel fabrication make it now possible to control hydrogel rheology and utilize computational results to generate optimized scaffold architectures.

CRISPRoffT: comprehensive database of CRISPR/Cas off-targets.

The CRISPR (clustered regularly interspaced short palindromic repeats)/Cas (CRISPR-associated protein) programmable nuclease system continues to evolve, with in vivo therapeutic gene editing increasingly applied in clinical settings. However, off-target effects remain a significant challenge, hindering its broader clinical application. To enhance the development of gene-editing therapies and the accuracy of prediction algorithms, we developed CRISPRoffT (https://ccsm.uth.edu/CRISPRoffT/). Users can access a comprehensive repository of off-target regions predicted and validated by a diverse range of technologies across various cell lines, Cas enzyme variants, engineered sgRNAs (single guide RNAs) and CRISPR editing systems. CRISPRoffT integrates results of off-target analysis from 74 studies, encompassing 29 experimental prediction techniques, 368 guide sequences, 226164 potential guide and off-target pairs and 8840 validated off-targets. CRISPRoffT features off-target data from different CRISPR approaches (knockout, base editing and prime editing) applied under diverse experimental conditions, including 85 different Cas/guide RNA (gRNA) combinations used across 34 different human and mouse cell lines. CRISPRoffT provides results of comparative analyses for individual guide sequences, genes, cell types, techniques and Cas/gRNA combinations under different conditions. CRISPRoffT is a unique resource providing valuable insights that facilitate the safety-driven design of CRISPR-based therapeutics, inform experimental design, advance the development of computational off-target prediction algorithms and guide RNA design algorithms.

Image modeling algorithm for environment design based on augmented and virtual reality technologies

Abstract With the continuous development of computer vision technology, the image modeling of environmental design has also received more and more attention. This study proposed an image modeling algorithm for environment design based on augmented reality (AR) and virtual reality (VR) technologies. On the system architecture, the environment design image modeling algorithm based on AR and VR technologies has provided users with an immersive 3D simulation space environment. It also has provided users with a good interactive interface through AR technology, enabling users to interact with computers in a realistic virtual three-dimensional environment and interactive operation. This study verified the algorithm from three aspects: time, authenticity, and interactivity. When facing object A, the modeling time values of the traditional environmental design image modeling method and the environmental design image modeling method proposed in this study were 12.14 and 8.72 s, respectively. The modeling authenticity reached 87.51 and 98.57%, respectively, and the interactivity reached 79.14 and 96.45%, respectively. Some columns of data have proved that the environmental design image modeling method proposed in this study was superior to the traditional environmental design image modeling method. In addition, this study also verified the system model from three aspects: stability, running smoothness, and performance loss. When the number of image modeling in the environment design was 2, the stability, running smoothness, and performance loss of the system were 99.37, 99.25 and 0.75%, respectively. The experimental data again proved that the algorithm system model proposed in this study was feasible and worthy of further application.

Recent Advances in controlled manipulation of micro/nano particles: A review

Abstract This review discusses the transformative impact of micro/nano particle manipulation techniques across scientific and technological disciplines. Emphasizing the pivotal role of precise control at the micro and nanoscale, the paper categorizes manipulation strategies into mechanical/surface force-based, field-control manipulation, and microfluidics manipulation. It addresses challenges specific to the submicrometer scale, highlighting the strengths and limitations of each approach. The unique behaviors exhibited by objects at the micro–nano scale influence the design and operation of manipulators, algorithms, and control systems, particularly in interactions with biological systems. The review covers dielectrophoresis and magnetic manipulation, showcasing their applications in particle manipulation and microfluidics. The evolution of optical tweezers, including holographic, surface plasmon-based, and optical fiber tweezers, is discussed, emphasizing their contributions in various scientific fields. Additionally, the paper also explores the manipulation of micro/nano particle in microfluidic platforms. The comprehensive review underscores the significance of understanding manipulation strategies in diverse environments, anticipating further advancements in science and technology.

A multi-objective energy-efficient scheduling algorithm for solving hybrid flow workshops with parallel heterospeed machines

Abstract. In hybrid flow assembly workshops, minimizing completion time and reducing energy consumption are two key indicators of efficient workshop management, which are often inversely related. This study investigates how parallel heterospeed machines with adjustable times can reduce energy costs and assembly times at the production system level. To address these issues, we propose the MO-HFSP-Batch algorithm based on the enhanced NSGA-II algorithm, aimed at improving the algorithm's optimization capability. Given the trade-offs between optimization objectives and the high computational complexity of the proposed multi-objective mixed integer programming, a three-tier chromosome encoding structure was introduced during the algorithm design phase. This structure meets the triple requirements of batching assembly tasks, matching operations with parallel machines, and sequencing operations in actual hybrid flow assembly workshops. Extensive data analysis proves that our proposed algorithm effectively solves the scheduling issues in hybrid flow workshops and performs better in solving completion time indicators and stability of non-dominated solution sets than other algorithms.

Merged-nets enumeration for the systematic design of multicomponent reticular structures.

Rational design of intricate multicomponent reticular structures is often hindered by the lack of suitable blueprint nets. We established the merged-net approach, proffering optimal balance between designability and complexity, as a systematic solution for the rational assembly of multicomponent structures. In this work, by methodically mapping node-net relationships among 53 basic edge-transitive nets, we conceived a signature net map to identify merging net pairs, resulting in the enumeration of 53 merged nets. We developed a practical design algorithm and proposed more than 100 multicomponent metal-organic framework platforms. The effectiveness of this approach is commended by the successful synthesis of four classes of materials, which is based on merging three-periodic nets with the four possible net periodicities. The construction of multicomponent materials based on derived nets of merged nets highlights the potential of the merged-net approach in accelerating the discovery of intricate reticular materials.

Critical Assessment of Protein Engineering (CAPE): A Student Challenge on the Cloud.

The success of AlphaFold in protein structure prediction highlights the power of data-driven approaches in scientific research. However, developing machine learning models to design and engineer proteins with desirable functions is hampered by limited access to high-quality data sets and experimental feedback. The Critical Assessment of Protein Engineering (CAPE) challenge addresses these issues through a student-focused competition, utilizing cloud computing and biofoundries to lower barriers to entry. CAPE serves as an open platform for community learning, where mutant data sets and design algorithms from past contestants help improve overall performance in subsequent rounds. Through two competition rounds, student participants collectively designed >1500 new mutant sequences, with the best-performing variants exhibiting catalytic activity up to 5-fold higher than the wild-type parent. We envision CAPE as a collaborative platform to engage young researchers and promote computational protein engineering.