1. Improved relation extraction through key phrase identification using community detection on dependency treesShuang Liu, Xunqin Chen, Jiana Meng, Niko Lukač, 2025, izvirni znanstveni članek Opis: A method for extracting relations from sentences by utilizing their dependency trees to identify key phrases is presented in this paper. Dependency trees are commonly used in natural language processing to represent the grammatical structure of a sentence, and this approach builds upon this representation to extract meaningful relations between phrases. Identifying key phrases is crucial in relation extraction as they often indicate the entities and actions involved in a relation. The method uses community detection algorithms on the dependency tree to identify groups of related words that form key phrases, such as subject-verb-object structures. The experiments on the Semeval-2010 task8 dataset and the TACRED dataset demonstrate that the proposed method outperforms existing baseline methods.
Ključne besede: community detection algorithms, dependency trees, relation extraction
Objavljeno v DKUM: 17.01.2025; Ogledov: 0; Prenosov: 1
 Celotno besedilo (3,12 MB) |
2. |
3. A review of federated learning in agricultureKrista Rizman Žalik, Mitja Žalik, 2023, pregledni znanstveni članek Opis: Federated learning (FL), with the aim of training machine learning models using data and computational resources on edge devices without sharing raw local data, is essential for improving agricultural management and smart agriculture. This study is a review of FL applications that address various agricultural problems. We compare the types of data partitioning and types of FL (horizontal partitioning and horizontal FL, vertical partitioning and vertical FL, and hybrid partitioning and transfer FL), architectures (centralized and decentralized), levels of federation (cross-device and cross-silo), and the use of aggregation algorithms in different reviewed approaches and applications of FL in agriculture. We also briefly review how the communication challenge is solved by different approaches. This work is useful for gaining an overview of the FL techniques used in agriculture and the progress made in this field.
Ključne besede: federated learning, agriculture, architecture, data partitioning, federation scal, aggregation algorithms, communication bottleneck
Objavljeno v DKUM: 05.06.2024; Ogledov: 142; Prenosov: 17
Celotno besedilo (839,33 KB)
Gradivo ima več datotek! Več... |
4. |
5. NiaNetAD: Autoencoder architecture search for tabular anomaly detection powered by HPCSašo Pavlič, Sašo Karakatič, Iztok Fister, 2023, objavljeni znanstveni prispevek na konferenci Ključne besede: predictive maintenance, autoencoder, anomaly detection, nature-inspired algorithms, optimization, high-performance computing, unsupervised learning
Objavljeno v DKUM: 26.01.2024; Ogledov: 447; Prenosov: 0 |
6. GeMMa Activity Report 2016-2022Tamara Golob, 2023, zbornik Opis: The primary aim of this publication is to support the dissemination of the research achievements of GeMMA Laboratory at the Faculty of Electrical Engineering and Computer Science of the University of Maribor. It follows the first survey published in 2016, and therefore, the actual book concentrates on the research results since then. The previous book, covering a period of 17 years, presented 35 R&D projects, while the new review of activities over the last 7 years covers as many as 58 projects. This growth is a good cue to introduce the secondary, equally important aim of the book. Namely, we would like to leave the track to our successors to stimulate them for even better research results, to show them, what is possible to achieve in 22 years starting from scratch with the will, hard work, orientation towards the applications, devotion to the research work, and the strong team spirit.
Ključne besede: research & development projects, geospatial modelling, multimedia, artificial intellingence, algorithms, industrial cooperation, international cooperation
Objavljeno v DKUM: 09.03.2023; Ogledov: 547; Prenosov: 76
Celotno besedilo (30,44 MB)
Gradivo ima več datotek! Več... |
7. Optimization methods for a direct current electric motor design : magistrsko deloVid Černec, 2022, magistrsko delo Opis: In the modern world, the need for planning in advance has become increasingly important. The companies want to know the cost, the dimensions of the elements, and the properties of the final product in advance. Therefore, we can find teams working on the analysis and development of new and different products in larger companies. More experienced workers can determine the given results and see what would be acceptable by simply looking at the situation. Although, the help of technology makes the process easier and faster, and even provides the same result. When creating a specific model, we must focus on the equations describing our result. Then we begin with optimization, which means we define the variables, which will be our subject of research in the project, and define some fixed parameters until the desired goals are reached.
Ključne besede: design, DC electric motor, optimization, optimization algorithms, genetic algorithms
Objavljeno v DKUM: 21.10.2022; Ogledov: 685; Prenosov: 53
Celotno besedilo (1,34 MB) |
8. Distance-based Invariants and Measures in GraphsAleksander Kelenc, 2019, doktorska disertacija Opis: This doctoral dissertation is concerned with aspects on distance related topics in graphs. We study three main topics, namely a recently introduced measure called the Hausdorff distance of graphs and two new graph invariants - the edge metric dimension and the mixed metric dimension of graphs. All three topics are part of the metric graph theory since they are tightly connected with the basic concept of distance between two vertices of a graph.
The Hausdorff distance is a relatively new measure of the similarity of graphs.
The notion of the Hausdorff distance considers a special kind of common subgraph of the compared graphs and depends on the structural properties outside of the common subgraph. We study the Hausdorff distance between certain families of graphs that often appear in chemical graph theory. Next to a few results for general graphs, we determine formulae for the distance between paths and cycles.
Previously, there was no known efficient algorithm for the problem of determining the Hausdorff distance between two trees, and in this dissertation we present a polynomial-time algorithm for it. The algorithm is recursive and it utilizes the divide and conquer technique. As a subtask it also uses a procedure that is based on the well-known graph algorithm for finding a maximum bipartite matching.
The edge metric dimension is a graph invariant that deals with distinguishing the edges of a graph. Let $G=(V(G),E(G))$ be a connected graph, let $w \in V(G)$ be a vertex, and let $e=uv \in E(G)$ be an edge. The distance between the vertex $w$ and the edge $e$ is given by $d_G(e,w)=\min\{d_G(u,w),d_G(v,w)\}$. A vertex $w \in V(G)$ distinguishes two edges $e_1,e_2 \in E(G)$ if $d_G(w,e_1) \ne d_G(w,e_2)$. A set $S$ of vertices in a connected graph $G$ is an edge metric generator of $G$ if every two distinct edges of $G$ are distinguished by some vertex of $S$. The smallest cardinality of an edge metric generator of $G$ is called the edge metric dimension and is denoted by $dim_e(G)$. The concept of the edge metric dimension is new. We study its mathematical properties. We make a comparison between the edge metric dimension and the standard metric dimension of graphs while presenting some realization results concerning the two. We prove that computing the edge metric dimension of connected graphs is NP-hard and give some approximation results. Moreover, we present bounds and closed formulae for the edge metric dimension of several classes of graphs.
The mixed metric dimension is a graph invariant similar to the edge metric dimension that deals with distinguishing the elements (vertices and edges) of a graph. A vertex $w \in V(G)$ distinguishes two elements of a graph $x,y \in E(G)\cup V(G)$ if $d_G(w,x) \ne d_G(w,y)$. A set $S$ of vertices in a connected graph $G$ is a mixed metric generator of $G$ if every two elements $x,y \in E(G) \cup V(G)$ of $G$, where $x \neq y$, are distinguished by some vertex of $S$. The smallest cardinality of a mixed metric generator of $G$ is called the mixed metric dimension and is denoted by $dim_m(G)$. In this dissertation, we consider the structure of mixed metric generators and characterize graphs for which the mixed metric dimension equals the trivial lower and upper bounds. We also give results on the mixed metric dimension of certain families of graphs and present an upper bound with respect to the girth of a graph. Finally, we prove that the problem of determining the mixed metric dimension of a graph is NP-hard in the general case.
Ključne besede: Hausdorff distance, distance between graphs, graph algorithms, trees, graph similarity, edge metric dimension, edge metric generator, mixed metric dimension, metric dimension
Objavljeno v DKUM: 03.08.2020; Ogledov: 1587; Prenosov: 124
Celotno besedilo (800,48 KB) |
9. Nature-inspired algorithms for hyperparameter optimization : magistrsko deloFilip Glojnarić, 2019, magistrsko delo Opis: This master thesis is focusing on the utilization of nature-inspired algorithms for hyperparameter optimization, how they work and how to use them. We present some existing methods for hyperparameter optimization as well as propose a novel method that is based on six different nature-inspired algorithms: Firefly algorithm, Grey Wolf Optimizer, Particle Swarm Optimization, Genetic algorithm, Differential Evolution, and Hybrid Bat algorithm. We also show the optimization results (set of hyperparameters) for each algorithm and we present the plots of the accuracy for each combination and handpicked one. In discussion of the results, we provide the answers on our research questions as well as propose ideas for future work.
Ključne besede: artificial intelligence, artificial neural networks, machine learning, nature-inspired algorithms, evolutionary algorithms
Objavljeno v DKUM: 09.12.2019; Ogledov: 2232; Prenosov: 122
Celotno besedilo (969,13 KB) |
10. Differential evolution and large-scale optimization applicationsAleš Zamuda, znanstveni film, znanstvena zvočna ali video publikacija Opis: Differential Evolution (DE) is one of the most popular, high-performance optimization algorithms with variants that have been outperforming others for years. As a result, DE has grown to accommodate wide usage for a variety of disciplines across scientific fields. Differential Evolution and Large-Scale Optimization Applications presents a research-based overview and cross-disciplinary applications of optimization algorithms. Emphasizing applications of Differential Evolution (DE) across sectors and laying the foundation for further use of DE algorithms in real-world settings, this video is an essential resource for researchers, engineers, and graduate-level students. Topics Covered : Algorithms, Optimization, Parallel Differential Evolution, Performance Improvement, Stochastic Methods, Tree Model Reconstruction.
Ključne besede: differential Evolution, optimization, algorithms, stochastic methods, tree models, tree model reconstruction
Objavljeno v DKUM: 14.05.2019; Ogledov: 1822; Prenosov: 224
 Povezava na datoteko
Gradivo ima več datotek! Več... |
