Reducing overfitting and improving generalization in training convolutional neural network under limited sample sizes in image recognition

in 5th International Conference on Information Technology, Bangsaen 21-22 October 2020 (2020)

Chaotic Traversal (CHAT): Very Large Graphs Traversal Using Chaotic Dynamics

in International Journal of Bifurcation and Chaos (2017), 27(14), 1750215

Graph Traversal algorithms can find their applications in various fields such as routing problems, natural language processing or even database querying. The exploration can be considered as a first ... [more ▼]

Graph Traversal algorithms can find their applications in various fields such as routing problems, natural language processing or even database querying. The exploration can be considered as a first stepping stone into knowledge extraction from the graph which is now a popular topic. Classical solutions such as Breadth First Search (BFS) and Depth First Search (DFS) require huge amounts of memory for exploring very large graphs. In this research, we present a novel memoryless graph traversal algorithm, Chaotic Traversal (CHAT) which integrates chaotic dynamics to traverse large unknown graphs via the Lozi map and the Rössler system. To compare various dynamics effects on our algorithm, we present an original way to perform the exploration of a parameter space using a bifurcation diagram with respect to the topological structure of attractors. The resulting algorithm is an efficient and nonresource demanding algorithm, and is therefore very suitable for partial traversal of very large and/or unknown environment graphs. CHAT performance using Lozi map is proven superior than the, commonly known, Random Walk, in terms of number of nodes visited (coverage percentage) and computation time where the environment is unknown and memory usage is restricted. [less ▲]

Metaheuristic Based Clustering Algorithms for Biological Hypergraphs

in Proceedings of META’2016, 6th International Conference on Metaheuristics and Nature Inspired computing (2016, October 27)

Hypergraphs are widely used for modeling and representing relationships between entities, one such field where their application is prolific is in bioinformatics. In the present era of big data, sizes and ... [more ▼]

Hypergraphs are widely used for modeling and representing relationships between entities, one such field where their application is prolific is in bioinformatics. In the present era of big data, sizes and complexity of these hypergraphs grow exponentially, it is impossible to process them manually or even visualize their interconnectivity superficially. A common approach to tackle their complexity is to cluster similar data nodes together in order to create a more comprehensible representation. This enables similarity discovery and hence, extract hidden knowledge within the hypergraphs. Several state-of-the-art algorithms have been proposed for partitioning and clustering of hypergraphs. Nevertheless, several issues remain unanswered, improvement to existing algorithms are possible, especially in scalability and clustering quality. This article presents a concise survey on hypergraph-clustering algorithms with the emphasis on knowledge-representation in systems biomedicine. It also suggests a novel approach to clustering quality by means of cluster-quality metrics which combines expert knowledge and measurable objective distances in existing biological ontology. [less ▲]