Search Space Reduction Research Articles

Enabling knowledge discovery through low utility itemset mining

While high utility mining aims to identify patterns that maximize profits, low utility mining (LUIM) focuses on uncovering exceptional patterns, enabling proactive interventions before issues escalate. Existing high utility itemset mining methods use upper-bound pruning to reduce the search space. However, this approach is ineffective for LUIM, as it inadvertently eliminates low utility itemsets (LUIs). In addition, LUIM struggles with distinguishing between LUIs and zero-utility candidates. These challenges in LUIM can be summarized as follows: (1) reducing the search space of LUIM, and (2) distinguishing genuine LUIs from zero-utility candidates. To address these issues, we propose LUIMiner, an algorithm designed to accurately and efficiently find the complete set of LUIs. LUIMiner incorporates two lower-bound pruning strategies—depth and width—to reduce search space and streamline the mining process. We also introduce a redesigned search tree that processes candidates from long to short, enabling effective lower-bound-based pruning. Additionally, a preprocessing phase filters out zero-utility itemsets early in the process by calculation of maximal non-mutually contained itemsets. The Bgenerator structure uses bitwise operations to accelerate computations and reduce database scans. Experimental comparisons with state-of-the-art LUG-Miner & LUIMA demonstrate the effectiveness and efficiency of our algorithm.

SinkFlow: Fast and traceable root-cause localization for multidimensional anomaly events

With the development of various artificial intelligence (AI)–based applications, detecting anomalies and analyzing the root causes from massive data are critical to increasing the usability of AI. Fast, accurate root-cause analysis (RCA) that finds the main reason for an anomaly, as well as reasonable explanations, helps in solving problems effectively. Thus, RCA plays an important role in troubleshooting and fault diagnosis, making its application in data analysis crucial. Previous root-cause-localization approaches for multidimensional anomaly events encompass various techniques to reduce search space and have improved the localization performance. However, they do not effectively balance the requirements in terms of performance, compatibility, and interpretability. To solve these problems, we propose a new root-cause-localization method called SinkFlow. It provides a unified framework event-aggregation Graph (EAG) to describe the constraints of event aggregation and relations between events, so it can be easily generalized to various domains. SinkFlow introduces an applicable measure evaluation method for both fundamental and derived measures to quantify the impact of events. Also, it utilizes an optimal search strategy to reduce the search space based on the anomaly behavioral consistency and deviation significance. Our experimental results on semisynthetic datasets show that SinkFlow achieved better performance than other baselines and ran much faster, achieving a 1.88% increase of the F1-score and only 25% of the time cost of the second best localization method. In addition, SinkFlow offered clear, visible explanations of the localization results to answer the questions of why they are root causes and how the anomaly is formed.

Accelerated bender’s decomposition algorithm and hybrid heuristics for multi-period planning of maternal healthcare facilities in India

This work addresses the challenge of improving availability and accessibility of maternal healthcare in India by presenting a multi-period planning problem of hierarchical and successively inclusive healthcare facilities. The problem is formulated as a mixed-integer linear programming model to minimize the overall cost, including the cost of establishing and upgrading the facilities, the cost of allocating/referring the mothers-to-be to the respective facilities and the penalty cost of demotivating the overburdening of the facilities in each time period. To solve the model effectively and efficiently, a Bender’s Decomposition Algorithm (BDA) with several acceleration strategies such as valid inequalities, disaggregated Benders cuts, rolling horizon heuristic and parallelism is developed. A Bender’s type heuristic is also tested by solving the master problem heuristically. Additionally, a Fix-and-Optimize (F&O) heuristic hybridized with Simulated Annealing (SA) enhanced by various search space reduction techniques is developed to obtain good quality solutions in a reasonable time for large instances. It is evident from the results of the computational experiments that the accelerated BDA and Bender type heuristic outperforms Gurobi. The hybrid F&O and SA is observed to be the most computationally efficient approach. A representative scenario in the Indian setting presents further evidence of the model’s applicability.

Redundancy resolution of a variable base frame of a 3-DoF cable-driven serial chain by using an adaptive neuro-fuzzy controller

This paper presents a novel approach using adaptive neuro-fuzzy techniques to design controllers for planar cable-driven serial chain robots with variable configurations. The approach consists of two key components: (1) deriving dynamic models for cable-driven serial chain robots which are independent of their structure, and (2) adaptively determining the optimal cable connection points. Traditional methods face challenges in obtaining accurate dynamic equations for cable-driven serial chain robots with high degrees-of-freedom, hence neural networks are employed to estimate the model. In order to handle the variability in cable connection points, adaptive fuzzy methods are utilized. The proposed adaptive neuro-fuzzy controller algorithm introduces two new indices, namely cost-of-redundancy and degree-of-redundancy, to effectively address redundancy concerns. Additionally, the algorithm efficiently reduces the search space for finding the optimal configuration. Simulation results for a planar 3 degrees-of-freedom cable-driven serial chain robot using this algorithm showcase a noteworthy 42% reduction in cost-of-redundancy and an impressive 53.125% reduction in search space.

Shapley-Guided Pruning for Efficient Graph Neural Architecture Prediction in Distributed Learning Environments

Graph neural architecture search methods have made significant strides in predictive modeling; however, challenges persist in optimizing these approaches for improved accuracy and scalability. In this paper, we propose GraphNAP++, an innovative framework that combines Shapley-value-guided search space pruning with data-centric distributed learning (DCDL) to address the scalability and efficiency limitations of existing GNAS methodologies. The proposed method begins by selecting a subset of architectures from the search space, which are evaluated on a graph validation dataset using DCDL. These architectures are encoded, and a neural predictor is trained to predict their performance. Shapley values are utilized to prune the search space by retaining only the most influential architectures. The neural predictor then estimates the performance of all architectures within the reduced search space, and the highest-performing architectures are selected for final evaluation. Experimental results on benchmark datasets in distributed learning environments demonstrate the effectiveness of GraphNAP++ in both graph and node classification tasks, highlighting its potential to advance the state-of-the-art in GNNs.

A Word-Level Adversarial Attack Method Based on Sememes and an Improved Quantum-Behaved Particle Swarm Optimization.

The goal of textual adversarial attack methods is to replace some words in an input text in order to make the victim model misbehave. This article proposes an effective word-level adversarial attack method based on sememes and an improved quantum-behaved particle swarm optimization (QPSO) algorithm. The sememe-based substitute method, which uses the words sharing the same sememes as the substitutes of the original words, is first employed to form the reduced search space. Then, an improved QPSO algorithm, called historical information-guided QPSO with random drift local attractor (HIQPSO-RD), is proposed to search the reduced search space for adversarial examples. The HIQPSO-RD introduces historical information into the current mean best position of the QPSO, for the purpose of improving the convergence speed of the algorithm, by enhancing its exploration ability and preventing the premature convergence of the swarm. The proposed algorithm uses the random drift local attractor technique to make a good balance between its exploration and exploitation, so that the algorithm can find a better adversarial attack example with low grammaticality and perplexity (PPL). In addition, it employs a two-stage diversity control strategy to enhance the search performance of the algorithm. Experiments on three natural language processing (NLP) datasets, with three commonly used nature language processing models as victim models, show that our method achieves higher attack success rates but lower modification rates than the state-of-the-art adversarial attack methods. Moreover, the results of human evaluations show that adversarial examples generated by our method can better maintain the semantic similarity and grammatical correctness of the original input.

NaCTR: Natural product-derived compound-based drug discovery pipeline from traditional oriental medicine by search space reduction

The drug discovery pipelines require enormous time and cost, albeit their infamously high risk of failures. Reducing such risk has therefore been the utmost goal in the process. Recently, natural products (NPs) in traditional oriental medicine (TOM) have come into the spotlight for their efficacy and safety supported throughout the history. Not only that, with the ever-increasing repository of various biological datasets, many data-driven in silico approaches have also been extensively studied for better efficient search and testing. However, TOM-based datasets lack information on recently prevalent diseases, while experimental datasets are prone to provide target spaces that are too large. Adequate combination of both approaches can therefore fill in each other's blanks. In this study, we introduce NaCTR, an in silico discovery pipeline that achieves such integration to suggest NPs-derived drug candidates for a given disease. First, phenotypes and disease genes for the disease are identified in literature and public databases. Secondly, a pool of potentially therapeutic NPs are identified based on both TOM-based phenotype records and compound-gene interaction datasets. Lastly, the compounds contained in the NPs are further screened for toxicity and pharmacokinetic properties. We use the Parkinson's disease as the case study to test the NaCTR pipeline. Through the pipeline, we propose glutathione and four other compounds as novel drug candidates. We further highlight the finding with literature support. As the first to effectively combine data from ancient and recent repositories, the NaCTR pipeline can be a novel pipeline that can be applied successfully to any other diseases.

A planning framework for intra-city electric bus road transportation systems

Gasoline-based vehicle transportation plays a crucial role in environmental pollution, which eventually leads to serious health issues in urban areas. The scope of large-scale deployment of personalized electric vehicles in the near future can lean towards such issues but can overburden local distribution grids (LDGs), and the entities are producing green energy. In this context, an electric bus road transportation system (EBRTS) can be seen as an effective and promising way, although it is capital-intensive. EBRTs can be successfully deployed by developing a well-tailored planning model that encompasses amicable solutions while considering the conflicting interests of electric utilities and investors/transporters.This paper proposes a planning model for intra-city EBRTS, especially for metropolitan cities, to extract tangible benefits in favour of electric utilities and private players. The proposed framework envisages a bus fleet charging model and financial model within the spatial, technical, and financial constraints yet preserves the interests of electric utilities and investors by providing tentative solutions with all information necessary from the investor’s perspective. The methodology is free from optimization as the realistic constraints inherently employ search space reduction. The simulation results on a real public transportation system reveal the proposed methodology’s simplicity, rationality, applicability, and flexibility.

High-traversability and efficient path optimization for deep-sea mining vehicles considering complex seabed environmental factors

In cobalt-rich crust regions, rapid and safe pathfinding is critical for the operation of Deep-Sea Mining Vehicles (DSMVs). However, existing pathfinding solutions rarely account for traversability predictions in complex mining environments and extensive map data often results in computational inefficiencies. This paper proposes a High-Traversability and Efficient Path Optimization (HTEPO) strategy designed to ensure safe navigation and rapid decision-making accounting for complex seabed environmental factors. Initially, we develop a traversability assessment model combining the Analytic Hierarchy Process (AHP) with the Fuzzy Comprehensive Evaluation (FCE) method. This model assigns traversability scores to map out high-risk and impassable areas effectively. We then introduce a novel map model that integrates these scores with Voronoi diagrams to reduce search space and improve computing efficiency, ensuring safer route selections. Further, based on traversability index and map model, an enhanced A∗ algorithm equipped with adaptive heuristic weights is utilized to swiftly identify high-traversability paths for the DSMVs. Finally, three case studies with different seabed mining environments validate the excellent performance of the proposed method over other major existing methods in terms of path safety and computational efficiency.

Time-efficient model predictive control for autonomous tugs with adaptive input constraints

The marine autonomous surface ship (MASS) has gained significant attention because of its wide application in waterborne transportation in recent years. The trajectory control of the MASS is crucial in ensuring safety, particularly in narrow navigation areas and urgent encounter situations. Model predictive control (MPC) is well known as a sufficient method to solve the tracking problem considering the actuator saturation with model uncertainties and environmental disturbances. However, due to the constraints on computing resources and hardware limits, the controller may fail to find a feasible solution for MPC within a reasonable amount of time, especially when the system models are coupled and complex. To improve the solution efficiency, this paper introduces adaptive constraints to reduce search space for optimization, i.e., the current tracking point’s speed is transformed into input constraints in the MPC formulation to decrease the computation burden, as well as reduce mechanical wear on the thrusters with smaller amplitudes of control actions. Simulations are carried out to evaluate the effectiveness of the proposed method with different MPC forms.

A survey of genetic algorithms for clustering: Taxonomy and empirical analysis

Clustering, an unsupervised learning technique, aims to group patterns into clusters where similar patterns are grouped together, while dissimilar ones are placed in different clusters. This task can present itself as a complex optimization problem due to the extensive search space generated by all potential data partitions. Genetic Algorithms (GAs) have emerged as efficient tools for addressing this task. Consequently, significant advancements and numerous proposals have been developed in this field.This work offers a comprehensive and critical review of state-of-the-art mono-objective Genetic Algorithms (GAs) for partitional clustering. From a more theoretical standpoint, it examines 22 well-known proposals in detail, covering their encoding strategies, objective functions, genetic operators, local search methods, and parent selection strategies. Based on this information, a specific taxonomy is proposed. In addition, from a more practical standpoint, a detailed experimental study is conducted to discern the advantages and disadvantages of approaches. Specifically, 22 different cluster validation indices are considered to compare the performance of clustering techniques. This evaluation is performed across 94 datasets encompassing diverse configurations, including the number of classes, separation between classes, and pattern dimensionality. Results reveal interesting findings, such as the key role of local search in optimizing results and reducing search space. Additionally, representations based on centroids and labels demonstrate greater efficiency and crossover and mutation operators do not prove to be as relevant. Ultimately, while the results are satisfactory, real-world clustering problems introduce additional complexity, especially for algorithms aiming to determine the number of clusters, resulting in diminished performance and the need for new approaches to be explored. Code, datasets and instructions to run algorithms in the LEAL library are available in an associated repository, in order to facilitate future experiments in this environment.

Search-Space Reduction via Essential Vertices

Search-Space Reduction via Essential Vertices

Predicting peptide properties from mass spectrometry data using deep attention-based multitask network and uncertainty quantification.

Database search algorithms reduce the number of potential candidate peptides against which scoring needs to be performed using a single (i.e. mass) property for filtering. While useful, filtering based on one property may lead to exclusion of non-abundant spectra and uncharacterized peptides - potentially exacerbating the streetlight effect. Here we present ProteoRift, a novel attention and multitask deep-network, which can predict multiple peptide properties (length, missed cleavages, and modification status) directly from spectra. We demonstrate that ProteoRift can predict these properties with up to 97% accuracy resulting in search-space reduction by more than 90%. As a result, our end-to-end pipeline is shown to exhibit 8x to 12x speedups with peptide deduction accuracy comparable to algorithmic techniques. We also formulate two uncertainty estimation metrics, which can distinguish between in-distribution and out-of-distribution data (ROC-AUC 0.99) and predict high-scoring mass spectra against correct peptide (ROC-AUC 0.94). These models and metrics are integrated in an end-to-end ML pipeline available at https://github.com/pcdslab/ProteoRift.

OPTIMIZING CARTON PRODUCT DELIVERY BY SOLVING TRAVELLING SALESMAN PROBLEM AT PACKAGING COMPANIES

Optimization of delivery routes is one form of increasing business productivity to achieve the company's main goal of distributing each product to customers. The Traveling Salesman Problem (TSP) is a combinatorial optimization problem where a salesman goes on a distribution journey that starts from the depot, then visits all customers exactly once, and ends with returning to the depot. This study aims to optimize the distribution route of carton products for packaging companies using the TSP model. The research methodology includes observation, interviews, and document studies to understand the distribution process of carton products at packaging companies. To complete the TSP model, Branch and Bound (BB) and Nearest Neighbor (NN) methods are applied to find the best solution for determining the distribution route of carton products. The way the BB method works is by utilizing branch cuts and boundaries to reduce search space and speed up the resolution process. In the NN method, the nearest point is chosen to get the shortest route distance. Research findings show that the use of the BB method results in a mileage difference from the initial route of 125 km (53% more efficient) and a fuel cost difference of 121,231 IDR (45% cheaper). Meanwhile, the NN method results in a mileage difference from the initial route of 115 km (48.94%) and a fuel cost difference of 109,901 IDR (41.27%). So, the method that produces the best solution is the BB method. The limitations of this study lie in the scale of the model used and the assumptions underlying the analysis. Future research can broaden the scope of the model and consider other factors that may affect the distribution of carton products. The results of this study contribute to improving the efficiency of carton product distribution.

Automated Identification of Ions Observed in Mass Spectra of Inorganic Compounds Using Isotopic Distribution Brute Force: Methodology and Performance Measurements.

This Article describes the method of isotopic distribution brute force, which can be used to identify ions registered in mass spectra of inorganic compounds in an automated manner when a library search cannot be conducted. A detailed description of the isotopic distribution brute force methodology is presented, including a discussion of computation-related difficulties. The ability of the proposed algorithm to identify various inorganic ions is tested on a small set of real-life low-resolution mass spectra of lead halides and copper halides. An evaluation of the isotopic distribution brute force performance is conducted using the low-resolution experimental mass spectra of natural rhenium sulfide and lead(II) chloride. Based on identification results and obtained performance measurements, we formulate the empirical restrictions on the input data, ensuring that the application of isotopic distribution brute force is feasible from the standpoints of search space reduction and identification time.

PC-SSRDE: A paradigm crossover-based differential evolution algorithm with search space reduction

The optimization of complex problems has always been a difficult task in the realm of evolutionary computation, as complex problems often have a large search space. Adding more dimensions to the decision variables also makes the search space more complicated, which slows down the algorithm. However, as the number of locally optimal solutions to the complex problem grows exponentially, it becomes easy for the algorithm to land in a local optimal region. In light of this background, this paper proposes a paradigm-crossover-based differential evolution algorithm with search space reduction and diversity exploration. During the evolution process, the proposed algorithm obtains the correlation coefficient for each dimension of the problem. Based on this correlation, it generates a paradigm that participates in crossover, accelerating the population’s movement towards promising regions.When the algorithm faces premature convergence and stagnation, it executes search space reduction and diversity exploration at the dimensional level to discard the unpromising search space and enhance the population diversity in the promising search space. We compared our proposed algorithm with eight state-of-the-art evolutionary algorithms in the CEC2017-BC test set to confirm its effectiveness, and the experimental results demonstrated its notable advantages for solution accuracy and convergence speed on high-dimensional complex problems.

A CNN pruning approach using constrained binary particle swarm optimization with a reduced search space for image classification

Deep convolutional neural networks (CNNs) have exhibited exceptional performance in a range of computer vision tasks. However, these deep CNNs typically demand significant computational resources, which not only hinders their practical deployment but also contributes to a considerable carbon footprint. To tackle this issue, several filter pruning methods based on evolutionary algorithms have been proposed to provide significant memory and energy savings during CNN inference. However, due to the curse of high dimensionality in the structure of deep CNNs, the search space expands dramatically, presenting significant challenges for these methods. This paper proposes a novel algorithm called BPSO-FPruner for CNN filter pruning. BPSO-FPruner utilizes a constrained binary particle swarm optimization algorithm for filter pruning, incorporating a new initialization strategy based on filter weighting information and a reduced search space strategy. Extensive validation using VGG, ResNet, DenseNet, and MobileNetv2 architectures on the CIFAR-10, CIFAR-100, and Tiny ImageNet datasets demonstrates the effectiveness of BPSO-FPruner in reducing model computational costs and carbon footprint emissions while maintaining or improving performance.

Roulette wheel-based level learning evolutionary algorithm for feature selection of high-dimensional data

Feature selection in high-dimensional data is a large-scale sparse and discrete optimization problem. Most evolutionary algorithms are designed to tackle continuous optimization problems. However, when dealing with high-dimensional feature selection tasks, they often suffer from poor population diversity and are computationally expensive. To address these challenges, this work introduces a roulette wheel-based level learning evolutionary algorithm (RWLLEA). RWLLEA integrates two key components. Firstly, it employs a leveled population mode. Individuals from higher levels provide guidance to those at lower levels during the evolutionary process, thereby exploring the potential combinatorial effects among features. Secondly, recognizing the characteristics associated with the high-dimensional feature selection task, a roulette wheel-based update method is devised to dynamically reduce search space and harmonize the algorithm's exploitation and exploration capacities across different stages. The performance of the proposed method was evaluated by comparing it with six other feature selection techniques across a range of fifteen diverse datasets. The experimental findings demonstrate that the proposed method can achieve a reduced feature set with a shorter runtime and exhibit superior classification accuracy.

Optimal pipe-sizing design of water distribution networks using modified Rao-II algorithm

ABSTRACT Several evolutionary algorithms (EAs) have been suggested in the last three decades for the least-cost design of water distribution networks (WDNs). EAs generally worked well to identify the global/near-global optimal solutions for small- to moderate-size networks in a reasonable time and computational effort. However, their applications to large-size networks are still challenging due to large computational effort. Recent developments in EAs are towards parameter-less techniques that avoid fine-tuning of case-specific parameters to reduce the computational effort. Further, several self-adaptive penalties and search-space reduction methodologies have been suggested to reduce the computational effort. A fast, efficient, and parameter-less Rao-II algorithm has been used earlier with penalty-based approaches for the optimal design of WDNs. In this study, the application of a Rao-II algorithm is further explored with three self-adaptive penalty approaches to compare the convergence characteristics. The Rao-II algorithm is observed to converge at an infeasible solution in cases that the applied penalty to an infeasible solution is not so substantial to make it inferior to the feasible solutions. Modifications are suggested to improve the Rao-II algorithm, named the modified Rao-II algorithm. The modified Rao-II algorithm with the self-adaptive penalty methods resulted in better solutions than those obtained earlier.

A self-training method based on fast binary bare-bones particle swarm optimization for semi-supervised classification

The generalization performance of self-training methods for semi-supervised classification depends on the selection of pseudo-labeled samples. While multiple variations of self-training methods are developed, strategies for selecting pseudo-labeled samples still suffer from the following challenging issues: a) heavily relying on specific assumptions about geometric and class relationships between different instances; b) parameter dependency; c) failure to effectively eliminate mislabeled pseudo-labeled noisy samples. These issues lead to low robustness and performance deterioration. To select pseudo-labeled samples more effectively and overcome the above-mentioned challenging issues, a self-training method based on fast binary bare-bones particle swarm optimization (ST-F3BPSO) is proposed. The core ideas of ST-F3BPSO are that a) a binary bare-bones particle swarm optimization (F3BPSO) with a new search space reduction strategy, a new particle update mechanism, and a new fitness function is first proposed; b) an F3BPSO-based sample sub-space optimization (SSO–F3BPSO) is second proposed to help ST-F3BPSO select pseudo-labeled samples and eliminate mislabeled pseudo-labeled noisy samples in the iterative self-training process without parameters and assumptions about geometric and class relationships of data. Intensive experiments on benchmark data sets have proven that a) ST-F3BPSO outperforms 7 state-of-the-art non-evolutionary algorithms-based self-training methods and 2 state-of-the-art evolutionary algorithms-based self-training methods in classification accuracy and marco F-measure of 2 popular base classifiers with various ratios of the initial labeled data or the class noise; b) ST-F3BPSO outperforms 2 state-of-the-art evolutionary algorithms-based self-training methods in computational efficiency.