Academia.eduAcademia.edu

Kruskal Algorithm

description19 papers
group5 followers
lightbulbAbout this topic
Kruskal's Algorithm is a greedy algorithm used in graph theory to find the minimum spanning tree of a connected, undirected graph. It operates by sorting the edges in ascending order of weight and adding them to the spanning tree, ensuring no cycles are formed, until all vertices are connected.
lightbulbAbout this topic
Kruskal's Algorithm is a greedy algorithm used in graph theory to find the minimum spanning tree of a connected, undirected graph. It operates by sorting the edges in ascending order of weight and adding them to the spanning tree, ensuring no cycles are formed, until all vertices are connected.

Key research themes

1. How do metaheuristic optimization algorithms improve clustering effectiveness compared to traditional methods like Kruskal's algorithm?

This research area investigates the application and enhancement of nature-inspired and metaheuristic algorithms to solve clustering problems, including NP-hard variants, outperforming classical algorithms such as Kruskal's MST algorithm. It matters as it addresses scalability, convergence, and solution quality challenges in high-dimensional and complex datasets where traditional approaches like Kruskal's face limitations.

Key finding: Introduced the Black Hole Algorithm (BH), a population-based metaheuristic inspired by black hole phenomena, which iteratively selects the best solution to attract others (stars), demonstrating superior performance over... Read more
Key finding: Developed a Multi-population Black Hole Algorithm (MBHA) which enhances exploration by relying on multiple best solutions instead of a single best, resulting in higher precision and robustness on nine benchmark functions and... Read more
Key finding: Evaluated two novel metaheuristics— Ions Motion Optimization and Weighted Superposition Attraction—showing competitive clustering performance against Particle Swarm Optimization and Artificial Bee Colony algorithms on... Read more
Key finding: Presented an enhanced clustering method integrating Differential Evolution (DE) with a clustering-specific chromosome encoding, outperforming K-means variants by employing advanced search and crossover strategies. By... Read more

2. What are the advanced initialization and parameter tuning techniques for K-Means that improve its convergence and clustering quality?

This research stream focuses on systematically improving K-Means through better initialization strategies, parameter control, and integration with clustering heuristics, aiming to alleviate K-Means’ sensitivity to initial seeds and convergence to local minima. Enhancements are essential since K-Means is NP-hard and prone to poor local optima, which limits its effectiveness especially in large, high-dimensional, or complex shaped data clusters where algorithms like Kruskal's MST indirectly inform clustering but lack adaptability.

Key finding: Conducted a systematic review of over a thousand documents to categorize K-Means improvements by algorithmic step (initialization, classification, centroid update, convergence), highlighting several highly cited variants.... Read more
Key finding: Proposed a novel geometry-based heuristic that determines the optimal number of clusters by analyzing the statistical distribution of an 'initial speed rate' parameter derived from data geometry. The heuristic identifies a... Read more

3. How can graph theory and combinatorial optimization methods like Kruskal's algorithm be integrated or adapted in metaheuristic frameworks to enhance clustering and networked system design?

This area explores the combination of classical graph algorithms (e.g., Kruskal's MST) with metaheuristic optimization to address complex problems such as clustering and network topology synthesis in dynamic systems. Leveraging graph-based representations and spanning tree solutions within or alongside metaheuristics unlocks richer exploitable structures for optimization, crucial where pure MST methods may yield suboptimal results or lack robustness against multimodal objective landscapes.

Key finding: Characterized the H2 norm of networked dynamic systems with homogeneous and heterogeneous agent dynamics in terms of the Frobenius norm of the graph incidence matrix and weighted variants respectively. By leveraging Kruskal’s... Read more
Key finding: Proposed a graph theory-based method to classify image regions into homogeneous or edges/detail areas and employed Kruskal's minimum spanning tree technique to facilitate smooth switching between filters for colour image... Read more
Key finding: Introduced a novel clustering operator integrating unsupervised k-windows clustering into evolutionary algorithms to identify multiple local and global minima in optimization problems. By grouping search individuals near... Read more

All papers in Kruskal Algorithm

Algoritma kruskal adalah salah satu algoritma yang digunakan untuk menyelesaikan masalah pencarian pohon rentang minimum (Minimum Spanning Tree, MST) pada graf berbobot tak terarah.
In 3 it was presented a way to i n troduce fuzzy edges and v ertices in graphs to m anipulate incertainty a n d imprecision in images. This procedure provide more exibility i n s e g m entation techniques based on pyramid strutures but i... more
Breve ensayo del algoritmo de Kruskal, pseudocódigo, ejemplo y complejidad.
In this study, the authors propose a soft-switching filter to improve the performance of recent colour image smoothing filters when processing homogeneous image regions. The authors use a recent filter mixed with the classical arithmetic... more
In this study, the authors propose a soft-switching filter to improve the performance of recent colour image smoothing filters when processing homogeneous image regions. The authors use a recent filter mixed with the classical arithmetic... more
Abstract: Let V nk denote the set of all n-dimensional vectors whose components are integers between 0 and k? 1. A subset of V nk is called a code. A code C is said to be an R-covering of V nk if, for all y 2 V nk, there exists a vector x... more
Recebido em 9 outubro 2018 / Aceito em 12 fevereiro 2019 RESUMO. Neste artigo é proposto um modelo de programac ¸ão por metas estendido aplicado ao planejamento de radioterapia, em que foi encontrada a melhor combinac ¸ão de pesos para as... more
This work consists in the application of an optimized breadth-first search (BFS) algorithm to select a couple of link-and-node-disjoint shortest-path between the two most remote users within an optical access network. Our results showed... more
Capagrafos c eset, SumarioC, cap1 i ntroducaoC, cap2 s ubgraf os, cap3 a rvores, cap4 d istancia, cap5 h am
Routing protocols present a latency, i.e. a time interval spent to up- date all routing tables, after the topology changes. During the convergence latency interval several packets and connections may be lost. This work pro- poses a new... more
Este trabalho tem por objetivo realizar um estudo sobre o problema de Alocação de Tutores em Aplicações de Provas nos polos do Cederj. Este problema consiste em encontrar uma alocação de disciplinas e tutores em salas de aula satisfazendo... more
Este trabalho tem por objetivo realizar um estudo sobre o problema de Alocação de Tutores em Aplicações de Provas nos polos do Cederj. Este problema consiste em encontrar uma alocação de disciplinas e tutores em salas de aula satisfazendo... more
O aumento crescente do número de veículos nos grandes centros urbanos tem acelerado o processo de deterioração das vias públicas, que não suportam tal fluxo. Esse problema é, em geral, solucionado parcialmente por operações chamadas de... more
Algoritmo de Dijkstra (arestas positivas).
Algoritmo de Bellman-Ford (caso geral).
Tipos de grafos. Busca em largura. Busca em profundidade.
O queé um grafo? Se você nunca ouviu falar nisso antes, estaé certamente uma pergunta que você deve estar se fazendo. Vamos tentar matar sua curiosidade contando como foi que a teoria dos grafos surgiu.
This work provides a general framework for the analysis and synthesis of a class of relative sensing networks (RSN) in the context of its H∞ performance. In an RSN, the underlying connection topology couples each agent at their outputs. A... more
This work provides a general framework for the analysis and synthesis of a class of linear networked dynamic systems (NDS). We focus our attention on NDS where the underlying connection topology couples the agents at their outputs. A... more
Download research papers for free!