Sökning: "Graph convolutional neural networks"

Visar resultat 1 - 5 av 8 uppsatser innehållade orden Graph convolutional neural networks.

1. 1. A Study of the Loss Landscape and Metastability in Graph Convolutional Neural Networks

   Master-uppsats, KTH/Matematisk statistik

   Författare :Sofia Larsson; [2020]
   Nyckelord :Graph neural networks; Graph convolutional neural networks; Loss landscape; Gradient descent; Stochastic gradient descent; Stochastic gradient Langevin dynamics; Grafneurala nätverk; grafiska faltningsnätverk; lösningslandskap; gradientmetoder; stokastiska gradientmetoder; stokastisk gradient Langevin dynamik;

   Sammanfattning : Many novel graph neural network models have reported an impressive performance on benchmark dataset, but the theory behind these networks is still being developed. In this thesis, we study the trajectory of Gradient descent (GD) and Stochastic gradient descent (SGD) in the loss landscape of Graph neural networks by replicating Xing et al. LÄS MER

3. 2. Real-time Anomaly Detection on Financial Data

   Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)

   Författare :Anna Martignano; [2020]
   Nyckelord :Network Representation Learning; Anomaly Detection; Financial Industry; Graph Neural Networks; Dynamic Graphs; Heterogeneous Graphs; Nätverkiskt Representationsinlärning; Avvikelse av Anomali; Finansiell Industri; Grafiskt Neuralt Nätverk; Dynamiska Grafer; Heterogena Grafer;

   Sammanfattning : This work presents an investigation of tailoring Network Representation Learning (NRL) for an application in the Financial Industry. NRL approaches are data-driven models that learn how to encode graph structures into low-dimensional vector spaces, which can be further exploited by downstream Machine Learning applications. LÄS MER

4. 3. Streaming Graph Partitioning with Graph Convolutional Networks

   Master-uppsats, KTH/Skolan för elektroteknik och datavetenskap (EECS)

   Författare :Michal Zwolak; [2020]
   Nyckelord :;

   Sammanfattning : In this work, we present a novel approach to the streaming graph partitioning problem which handles unbounded streams.Graph partitioning is a process of dividing a graph into groups of nodes or edges. LÄS MER

5. 4. Manipulation Action Recognition and Reconstruction using a Deep Scene Graph Network

   Master-uppsats, Högskolan i Halmstad/Akademin för informationsteknologi; Högskolan i Halmstad/Akademin för informationsteknologi

   Författare :Dawid Ejdeholm; Jacob Harsten; [2020]
   Nyckelord :;

   Sammanfattning : Convolutional neural networks have been successfully used in action recognition but are usually restricted to operate on Euclidean data, such as images. In recent years there has been an increase in research devoted towards finding a generalized model operating on non-Euclidean data (e. LÄS MER

7. 5. Imaging Using Machine Learning for the LDMX Electromagnetic Calorimeter

   Master-uppsats, Lunds universitet/Partikelfysik; Lunds universitet/Fysiska institutionen

   Författare :Leo Östman; [2020]
   Nyckelord :Physics and Astronomy;

   Sammanfattning : LDMX is a fixed target experiment designed to search for light dark matter. The experiment will search for dark matter signatures using missing momentum and energy in events with electrons scattering in the target. LÄS MER