Convolutional Signal Propagation: A Simple Scalable Algorithm for Hypergraphs
A simple, scalable and efficient baseline algorithm for classification and retrieval in hypergraphs, based on non-parametric convolution.
Balancing performance and complexity with adaptive graph coarsening
A method for node classification that allows a user to precisely select the resolution at which the graph in question should be pretrained.
Which Graph Properties Affect GNN Performance for a Given Downstream Task?
A methodical way of linking graph properties with the performance of a GNN solving a given task on such graph via a surrogate regression model that is trained to predict the performance of the GNN from the properties of the graph dataset.
Scalable Graph Size Reduction for Efficient GNN Application
A simple scalable task-aware graph preprocessing procedure allowing us to obtain a reduced graph such that a GNN achieves a given desired performance on a given downstream task.
Adaptive graph coarsening in the context of local graph quality
A method for studying graph properties from the point of view of a downstream task.
Experimental Investigation of Neural and Weisfeiler-Lehman-Kernel Graph Representations for Downstream Classification
An experimental comparison of four variants of 1-GNN and GIN from the point of view of graph representation for downstream classification.
Loss Functions for Clustering in Multi-instance Learning
Multi-instance clustering using contrastive predictive coding, triplet and magnet loss.
Nested Multiple Instance Learning in Modelling of HTTP network traffic
Identification of infected computers in the computer network from their HTTP traffic using recent progress in multiple-instance learning.
Optimalization of distances for multi-instance clustering
Multi-instance clustering using contrastive predictive coding, triplet and magnet loss.
Hierarchické modely síťového provozu
Automatický klasifikátor rozpoznávající aktivity malware na úrovni síťových spojení za pomoci multi-instančního učení.
