K-means Based Clustering Algorithm Research Articles

Multi-Cell-to-Multi-Cell active equalization method based on k-means clustering and battery pack SOC estimation

In this paper, a MC2MC (Multi-Cell-to-Multi-Cell) battery pack active equalization scheme based on reconfigurable circuits and Buck-Boost converters is proposed and validated for the problem of inconsistent single cell power in battery packs. This scheme can simultaneously equalize multiple non-adjacent cells in a single equalization process, improving the equalization time efficiency. Then, based on the relationship between duty cycle and voltage difference, the MC2MC battery pack active equalization control strategy based on adaptive duty cycle is designed. In addition, a k-means based clustering algorithm is used to cluster and reorganize batteries with different charges, and the SOC (State of charge) estimation algorithm of the battery pack is improved and introduced into the equalization control among multiple batteries. Finally, AC2AC (Any-Cell-to-any -Cell), MC2MC, and MC2MC simulations considering the SOC estimation of the battery pack are validated and analyzed for the unbalanced battery packs under MATLAB/Simulink environment, respectively. On the basis of verifying the accuracy of the simulation process, the effectiveness of the MC2MC equalization control method considering the SOC estimation of the battery pack is verified through the comparative analysis of four aspects, namely, equalization time, computational workload, resultant variance, and energy efficiency.

Machine learning-driven task scheduling with dynamic K-means based clustering algorithm using fuzzy logic in FOG environment

Fog Computing has emerged as a pivotal technology for enabling low-latency, context-aware, and efficient computing at the edge of the network. Effective task scheduling plays a vital role in optimizing the performance of fog computing systems. Traditional task scheduling algorithms, primarily designed for centralized cloud environments, often fail to cater to the dynamic, heterogeneous, and resource-constrained nature of Fog nodes. To overcome these limitations, we introduce a sophisticated machine learning-driven methodology that adapts task allocation to the ever-changing Fog environment's conditions. Our approach amalgamates K-Means clustering algorithm enhanced with fuzzy logic, a robust unsupervised learning technique, to efficiently group Fog nodes based on their resource characteristics and workload patterns. The proposed method combines the clustering capabilities of K-means with the adaptability of fuzzy logic to dynamically allocate tasks to fog nodes. By leveraging machine learning techniques, we demonstrate how tasks can be intelligently allocated to fog nodes, resulting in reducing execution time, response time and network usage. Through extensive experiments, we showcase the effectiveness and adaptability of our proposed approach in dynamic fog environments. Clustering proves to be a time-effective method for identifying groups of jobs per virtual machine (VM) efficiently. To model and evaluate our proposed approach, we have utilized iFogSim. The simulation results affirm the effectiveness of our scheduling technique, showcasing significant enhancements in execution time reduction, minimized network utilization, and improved response time when compared to existing machine learning and non-machine learning based scheduling methods within the iFogSim framework.

Multi-Agent Reinforcement Learning in NOMA-Aided UAV Networks for Cellular Offloading

A novel framework is proposed for cellular offloading with the aid of multiple unmanned aerial vehicles (UAVs), while non-orthogonal multiple access (NOMA) technique is employed at each UAV to further improve the spectrum efficiency of the wireless network. The optimization problem of joint three-dimensional (3D) trajectory design and power allocation is formulated for maximizing the throughput. Since ground mobile users are considered as roaming continuously, the UAVs need to be re-deployed timely based on the movement of users. In an effort to solve this pertinent dynamic problem, a K-means based clustering algorithm is first adopted for periodically partitioning users. Afterward, a mutual deep Q-network (MDQN) algorithm is proposed to jointly determine the optimal 3D trajectory and power allocation of UAVs. In contrast to the conventional deep Q-network (DQN) algorithm, the MDQN algorithm enables the experience of multi-agent to be input into a shared neural network to shorten the training time with the assistance of state abstraction. Numerical results demonstrate that: 1) the proposed MDQN algorithm is capable of converging under minor constraints and has a faster convergence rate than the conventional DQN algorithm in the multi-agent case; 2) The achievable sum rate of the NOMA enhanced UAV network is 23% superior to the case of orthogonal multiple access (OMA); 3) By designing the optimal 3D trajectory of UAVs with the MDON algorithm, the sum rate of the network enjoys 142% and 56% gains than invoking the circular trajectory and the 2D trajectory, respectively.

Bibliometric Study on the Use of Machine Learning as Resolution Technique for Facility Layout Problems

Abstract Facility Layout Problems (FLP) are concerned with finding efficient factory layouts. Numerous resolution approaches are known in literature for layout optimization. Among those, intelligent approaches are less researched than solutions from exact or approximating approaches. The recent surge of research interest in Artificial Intelligence, and specifically Machine Learning (ML) techniques, presages an increase of such techniques’ usage in FLP. However, previous reviews on FLP research induce that, to date, this trend has not yet emerged. Utilizing a systematic literature review coupled with a k-Means based clustering algorithm, we analyzed 25 relevant publication full-texts from an original sample of 1,425 papers. Our findings corroborate the statement that ML techniques have attracted substantially less research interest than most other resolution approaches. While a few papers used Unsupervised Learning algorithms directly as a solution to the FLP, Supervised and Reinforcement Learning were found to be practically irrelevant. ML usage was significantly higher in FLP-adjacent planning tasks such as group technology. Drawing from experiences with other NP-hard combinatorial optimization problems in manufacturing research, we conclude that Reinforcement Learning is most promising to bridge the evident gap between FLP and ML research. Our study further contributes to FLP research by extending established classification frameworks.

K-Means Clustering With Incomplete Data

Clustering has been intensively studied in machine learning and data mining communities. Although demonstrating promising performance in various applications, most of the existing clustering algorithms cannot efficiently handle clustering tasks with incomplete features which is common in practical applications. To address this issue, we propose a novel K-means based clustering algorithm which unifies the clustering and imputation into one single objective function. It makes these two processes be negotiable with each other to achieve optimality. Furthermore, we design an alternate optimization algorithm to solve the resultant optimization problem and theoretically prove its convergence. The comprehensive experimental study has been conducted on nine UCI benchmark datasets and real-world applications to evaluate the performance of the proposed algorithm, and the experimental results have clearly demonstrated the effectiveness of our algorithm which outperforms several commonly-used methods for incomplete data clustering.

A K-Means based clustering algorithm with balanced constraints

聚类是一种重要数据分析技术，在众多领域中得到广泛地应用.然而，由于数据分布的内在特点，传统的聚类算法并不能保证聚类结果具有平衡性，这与很多现实的需求不一致.本文提出了一种基于K-Means的平衡约束聚类算法，该算法对K-Means算法每次迭代中数据点的分配策略进行修改，达到对每个簇可包含的数据点数目上限进行约束的目的.同时，算法支持用户自定义簇可包含的数据点数目上限，满足不同的平衡约束聚类需求.另外，本算法参数少，只需设置目标簇数目及其可包含的数据点数目上限，时间复杂度低，具有简单、快速的特点.在6个UCI（University of CaliforniaIrvine）真实数据集上进行的实验结果表明，文中提出的平衡约束聚类算法相比其他平衡约束聚类算法具有更佳的聚类效果和时间性能.

A K-partitioning algorithm for clustering large-scale spatio-textual data

The volume of spatio-textual data is drastically increasing in these days, and this makes more and more essential to process such a large-scale spatio-textual dataset. Even though numerous works have been studied for answering various kinds of spatio-textual queries, the analyzing method for spatio-textual data has rarely been considered so far. Motivated by this, this paper proposes a k-means based clustering algorithm specialized for a massive spatio-textual data. One of the strong points of the k-means algorithm lies in its efficiency and scalability, implying that it is appropriate for a large-scale data. However, it is challenging to apply the normal k-means algorithm to spatio-textual data, since each spatio-textual object has non-numeric attributes, that is, textual dimension, as well as numeric attributes, that is, spatial dimension. We address this problem by using the expected distance between a random pair of objects rather than constructing actual centroid of each cluster. Based on our experimental results, we show that the clustering quality of our algorithm is comparable to those of other k-partitioning algorithms that can process spatio-textual data, and its efficiency is superior to those competitors.

A Survey on Feature Weighting Based K-Means Algorithms

In a real-world data set, there is always the possibility, rather high in our opinion, that different features may have different degrees of relevance. Most machine learning algorithms deal with this fact by either selecting or deselecting features in the data preprocessing phase. However, we maintain that even among relevant features there may be different degrees of relevance, and this should be taken into account during the clustering process. With over 50 years of history, K-Means is arguably the most popular partitional clustering algorithm there is. The first K-Means based clustering algorithm to compute feature weights was designed just over 30 years ago. Various such algorithms have been designed since but there has not been, to our knowledge, a survey integrating empirical evidence of cluster recovery ability, common flaws, and possible directions for future research. This paper elaborates on the concept of feature weighting and addresses these issues by critically analyzing some of the most popular, or innovative, feature weighting mechanisms based in K-Means.