Multiobjective Binary Differential Approach with Parameter Tuning for Discovering Business Process Models: MoD‐ProM

The Scientific World JOURNAL, Год журнала: 2024, Номер 2024(1)

Опубликована: Янв. 1, 2024

Process discovery approaches analyze the business data to automatically uncover structured information, known as a process model. The quality of model is measured using dimensions, completeness (replay fitness), preciseness, simplicity, and generalization. Traditional algorithms usually output single A may not accurately capture observed behavior overfit training data. We have formed problem in multiobjective framework that yields several candidate solutions for end user who can pick suitable based on local environmental constraints (possibly varying). consider Binary Differential Evolution approach task discovery. proposed method employs dichotomous crossover/mutation operators. parameters are tuned grey relational analysis combined with Taguchi approach. compared well‐known single‐objective state‐of‐the‐art evolutionary algorithm—Nondominated Sorting Genetic Algorithm (NSGA‐II). Additional comparison via computing weighted average dimensions also undertaken. Results show algorithm computationally efficient produces diversified score high fitness functions. It shown models generated by superior or at least good those algorithms.

Язык: Английский

---

Feature Selection Based on Improved White Shark Optimizer

Journal of Bionic Engineering, Год журнала: 2024, Номер unknown

Опубликована: Сен. 26, 2024

Язык: Английский

---

ACGRIME: adaptive chaotic Gaussian RIME optimizer for global optimization and feature selection

Cluster Computing, Год журнала: 2024, Номер 28(1)

Опубликована: Ноя. 4, 2024

Язык: Английский

---

Distributed sparsity constrained optimization over the Stiefel manifold

Neurocomputing, Год журнала: 2024, Номер 602, С. 128267 - 128267

Опубликована: Июль 26, 2024

Язык: Английский

---

Machine Learning Based Impairment-Aware Dynamic RMSCA in Multi-Core Elastic Optical Networks

Journal of Optical Communications and Networking, Год журнала: 2024, Номер 16(10), С. F26 - F26

Опубликована: Авг. 5, 2024

This paper presents a routing, modulation, spectrum, and core assignment (RMSCA) algorithm for space-division-multiplexing-based elastic optical networks (SDM-EONs) comprising multi-core links. A network state-dependent route selection method is proposed using deep neural (DNN) classifier. The DNN trained metaheuristic optimization to predict lightpath suitability, considering the quality of transmission resource availability. Physical layer impairments, including inter-core crosstalk, amplified spontaneous emission, Kerr fiber nonlinearities, are considered, random forest (RF)-based link noise estimator proposed. feature importance analysis provided all features considered classifier RF estimator. machine-learning-enabled RMSCA approach evaluated on three topologies, USNET, NSFNET, COST-239 with 7-core 12-core It shown be superior in terms blocking probability, bandwidth acceptable computational speed compared standard published benchmarks at different traffic loads.

Язык: Английский

---

Disaster tweet classification using enhanced salp swarm algorithm

Web Intelligence, Год журнала: 2024, Номер unknown, С. 1 - 17

Опубликована: Сен. 9, 2024

Twitter and Facebook are widely recognized as crucial tools for situational information during disasters. Given that the classification of disaster related tweets is computationally challenging due to high dimension textual data caused by redundant irrelevant features. Hence optimal feature selection (FS) tweets, this work utilizes binary salp swarm algorithm (BSSA) proposed two enhancements over it (PBcSSA). The commensalism phase from symbiotic organisms search (SOS) integrated with BSSA enhance its space searchability then parallel implementation done using Apache Spark framework reduce execution time. experiments were performed in a cross-disaster setting on nine groups datasets including biological, earthquake, flood, hurricane, industrial, societal, transportation, wildfire, environmental. PBcSSA combined Naive Bayes (NB) classifier wrapper mode performance compared standard BSSA, sine cosine (BSCA), particle optimization (BPSO), grey wolf (BGWO), whale (BWOA). experimental results reveal outperforms other algorithms tweet achieved highest average F1-score lowest set reduced

Язык: Английский

---

An evolutionary algorithm‐based classification method for high‐dimensional imbalanced mixed data with missing information

Electronics Letters, Год журнала: 2024, Номер 60(20)

Опубликована: Окт. 1, 2024

Abstract The data scale keeps growing by leaps and the majority of it is high‐dimensional imbalanced data, which hard to classify. Data missing often happens in software further aggravates difficulty classifying data. In order resolve mixed‐variables classification problem, a novel method based on particle swarm optimization developed. It has three original components including multiple feature selection, mixed attribute imputation, quantum oversampling. Multiple selection uses two‐stage strategy obtain stable relevant features. Mixed imputation separates features into continuous discrete fills values with different models. Quantum oversampling chooses instances balance operator. Furthermore, employed optimize parameters preferable results. Six representative datasets, six typical algorithms, four measures are taken conduct exhaust experiments, results indicate that proposed superior comparison algorithms.

Язык: Английский

---

A hybrid metaheuristic algorithm for antimicrobial peptide toxicity prediction

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Ноя. 16, 2024

The development of new algorithms can aid researchers and professionals in resolving problems that were once unsolvable or discovering superior solutions to already settled. By recognizing the importance continuous research on creating novel algorithms, this paper introduced a hybrid metaheuristic algorithm-h-PSOGNDO, which is combination Particle Swarm Optimization (PSO) Generalized Normal Distribution (GNDO). proposed algorithm utilizes Optimization's strategy for exploitation global search exploration. Through combination, h-PSOGNDO believed be an effective promote advantages its parents' algorithms. Different assessment methods are used assess algorithm. First, set conduct experiments two sets mathematical functions, including twenty-eight IEEE CEC2017 ten CEC2019 benchmark test respectively. Then, applied case study prediction antimicrobial peptides' toxicity evaluate performance real-life problems. statistical findings collected from both function show works effectively, proving astonishing ability yield highly competitive outcomes complex

Язык: Английский

---

A solution to common fixed point problems using a hybrid method of invasive weed optimization and jaya algorithm

Soft Computing, Год журнала: 2024, Номер unknown

Опубликована: Дек. 3, 2024

Язык: Английский

---