the proceedings contain 35 papers. the topics discussed include: a column generation heuristic for the general vehicle routing problem;a combination of evolutionary algorithm, mathematical programming, and a new local...
ISBN:
(纸本)3642137997
the proceedings contain 35 papers. the topics discussed include: a column generation heuristic for the general vehicle routing problem;a combination of evolutionary algorithm, mathematical programming, and a new local search procedure for the just-in-time job-shop scheduling problem;a math-heuristic algorithm for the DNA sequencing problem;a randomized iterated greedy algorithm for the founder sequence reconstruction problem;algorithm selection as a bandit problem with unbounded losses;bandit-based estimation of distribution algorithms for noisy optimization: rigorous runtime analysis;distance functions, clustering algorithms and microarray data analysis;Gaussian process assisted particle swarm optimization;learning of highly-filtered data manifold using spectral methods;multiclass visual classifier based on bipartite graph representation of decision tables;and a linear approximation of the value function of an approximate dynamic programming approach for the ship scheduling problem.
Simplifying the deployment, management, and maintenance of gadget mastering fashions within the manufacturing environment is a key aim in the swiftly growing system gaining knowledge of environment (MLOps). Intelligen...
详细信息
this paper introduces the Susceptible-Infected-Removed Optimizer (SIRO), a novel learned heuristic inspired by biological systems and deep learning techniques. SIRO models its search process after the SIR epidemiologi...
详细信息
this paper introduces the Susceptible-Infected-Removed Optimizer (SIRO), a novel learned heuristic inspired by biological systems and deep learning techniques. SIRO models its search process after the SIR epidemiological compartmental model, predicting the susceptibility, infection, and recovery dynamics of solutions. SIRO integrates deep learning into its initialization and parameter setting to enhance its efficiency, enabling intelligent and adaptive behavior. this hybridization improves solution quality, accelerates convergence, enhances robustness, and reduces computational costs. the algorithm's performance was evaluated using CEC 2017 benchmark functions, demonstrating superior results in hybrid functions (C1-C28) despite moderate performance on traditional CEC1-CEC14 functions. Friedman's test ranked SIRO 4th overall, with SSA as the top-performing algorithm. Additionally, SIRO was tested on real-world optimization problems, including mechanical engineering design, hyperparameter tuning, and feature selection for medical image classification. In the classification task, SIRO-enhanced CNN achieved an accuracy of 0.86 at the 5th epoch, outperforming CNN (0.66), CNN-GA (0.76), and CNN-WOA (0.75). Furthermore, SIRO reported a precision of 0.96, recall of 1.0, and F1-score of 0.98, highlighting its effectiveness. these results validate the benefits of integrating a learning mechanism into SIRO, yielding superior precision, computational efficiency, and performance over conventional optimization approaches.
In intelligent systems, as the amount of data increases, how to analyse data and extract abnormal information is an important task. Based on this, in order to improve the detection efficiency and accuracy of abnormal ...
详细信息
All machine learning procedures consume a mathematical foundation. the aforementioned is applicable to Deep learning, optimization, and any additional Statistics Science processes since Deep Knowledge is a subset of M...
详细信息
the smart grid is an advanced power system network that integrates cutting-edge technologies to enable efficient, reliable, and sustainable energy generation, distribution, and utilization. this paper provides a compr...
详细信息
the burgeoning demand for sustainable energy sources has catalyzed interest in Waste-to-Energy solutions offering both environmental benefits and energy generation potential. this paper explores the integration of mac...
详细信息
ISBN:
(纸本)9783031837951;9783031837968
the burgeoning demand for sustainable energy sources has catalyzed interest in Waste-to-Energy solutions offering both environmental benefits and energy generation potential. this paper explores the integration of machine learning (ML) techniques to optimize and predict key aspects of the Waste-to-Energy process. the study leverages historical data from waste facilities incorporating variables such as waste composition, operational parameters, and environmental conditions. In this paper, machine learning models, including regression, classification, and ensemble methods, are employed to optimize combustion efficiency, predict energy output, and enhance the overall operational performance of Waste-to-Energy conversion process. Furthermore, the predictive analytics are employed to anticipate maintenance needs for mitigating the downtime and optimizing the resource allocation. the findings contribute to the growing field of the sustainable energy by showcasing the efficacy of machine learning in Waste-to-Energy systems providing a scalable and adaptive solution for the challenges inherent in this dynamic and complex process.
Wireless Sensor Networks (WSNs) find extensive applications in environmental monitoring, healthcare, and smart cities. Energy efficiency, however, continues to be a significant challenge withthe limited lifetime of s...
详细信息
Accurate seasonal air quality prediction is crucial for effective environmental management and public health. this study introduces a novel approach that integrates Polynomial Regression with Bayesian optimization for...
详细信息
ISBN:
(纸本)9783031837920;9783031837937
Accurate seasonal air quality prediction is crucial for effective environmental management and public health. this study introduces a novel approach that integrates Polynomial Regression with Bayesian optimization for forecasting air quality indices (AQI). Traditional ensemble models such as Gradient Boosting, Random Forest, XGBoost, LightGBM and CatBoost, while effective, often face challenges like overfitting, high computational demands and complex hyperparameter tuning. Our approach addresses these issues by using Polynomial Regression to capture non-linear relationships and Bayesian optimization to streamline hyperparameter tuning, enhancing model efficiency and interpretability. the proposed model outperforms conventional methods based on Mean Squared Error (MSE). Seasonal pollutant analysis also highlights unique patterns in air quality changes, providing deeper insights for better environmental decision-making. Future research will focus on integrating advanced techniques and more variables to further refine and broaden the model's applicability.
the challenges of cost control in construction and installation projects have perennially been a significant concern for entities in the construction sector. the intricate interplay of various equipment, personnel, an...
详细信息
暂无评论