This paper is concerned with the impact of hubness, a general problem of machine learning in high-dimensional spaces, on a real-world music recommendation system based on visualisation of a k-nearest neighbour (knn) graph. Due to a problem of measuring distances in high dimensions, hub objects are recommended over and over again while anti-hubs are nonexistent in recommendation lists, resulting in poor reachability of the music catalogue. We present mutual proximity graphs, which are an alternative to knn and mutual knn graphs, and are able to avoid hub vertices having abnormally high connectivity. We show that mutual proximity graphs yield much better graph connectivity resulting in improved reachability compared to knn graphs, mutual knn graphs and mutual knn graphs enhanced with minimum spanning trees, while simultaneously reducing the negative effects of hubness.
Download full-text PDF |
Source |
|---|---|
| http://www.ncbi.nlm.nih.gov/pmc/articles/PMC5750815 | PMC |
| http://dx.doi.org/10.1080/09298215.2017.1354891 | DOI Listing |
Publication Analysis
Top Keywords
Similar Publications
Data-model interactive Rul prediction of stochastic degradation devices with multiple uncertainty quantification and multi-sensor information fusion.
ISA Trans
December 2024
College of Information Engineering, Zhejiang University of Technology, Hangzhou, 310023, China.
This paper proposes an improved remaining useful life (RUL) prediction method for stochastic degradation devices monitored by multi-source sensors under data-model interactive framework. Firstly, the interrelationships among sensors are established using k-nearest neighbor (KNN), and the composite health index (CHI) is constructed by aggregating the multi-source sensor information through the graph convolutional network (GCN). Secondly, a stochastic degradation model with triple uncertainty at any initial degradation level is established to improve the matching degree between the stochastic degradation model and the actual degradation process.
View Article and Find Full Text PDFEEG-based deception detection using weighted dual perspective visibility graph analysis.
Cogn Neurodyn
December 2024
Department of Biomedical Engineering, Amirkabir University of Technology (Tehran Polytechnic), Tehran, Iran.
Deception detection is a critical aspect across various domains. Integrating advanced signal processing techniques, particularly in neuroscientific studies, has opened new avenues for exploring deception at a deeper level. This study uses electroencephalogram (EEG) signals from a balanced cohort of 22 participants, consisting of both males and females, aged between 22 and 29, engaged in a visual task for instructed deception.
View Article and Find Full Text PDFConnectomics is a subfield of neuroscience that aims to map the brain's intricate wiring diagram. Accurate neuron segmentation from microscopy volumes is essential for automating connectome reconstruction. However, current state-of-the-art algorithms use image-based convolutional neural networks that are limited to local neuron shape context.
View Article and Find Full Text PDFPredicting variable-length ACE inhibitory peptides based on graph convolutional network.
Int J Biol Macromol
December 2024
Institute of Food Science and Technology, Chinese Academy of Agricultural Sciences/Key Laboratory of Agro-Products Processing, Ministry of Agriculture, Beijing 100193, China. Electronic address:
Traditional molecular descriptors have contributed to the prediction of angiotensin I-converting enzyme (ACE) inhibitory peptides, but they often fall short in capturing the complex structure of the molecule. To address these limitations, this study introduces molecular graphs as an advanced method for peptide characterization. Peptides containing 2-10 amino acids were represented using molecular graphs, and a graph convolutional network (GCN) model was constructed to predict variable-length peptides.
View Article and Find Full Text PDFA systematic review on machine learning approaches in cerebral palsy research.
PeerJ
October 2024
Electrical and Computer Engineering Department, University of Memphis, Memphis, TN, United States.
Article Synopsis
- This review investigates how machine learning (ML) models are being used to improve the understanding and management of cerebral palsy (CP), with a focus on identifying CP, classifying its subtypes, predicting abnormalities, and enhancing treatment strategies.
- A total of 20 relevant studies from 2013 to 2023 were analyzed, sourced from various academic databases like PubMed and IEEE Xplore, while ensuring they met specific inclusion criteria such as being peer-reviewed and related to ML applications for CP.
- The review highlights the importance of appropriate methodologies and robust study designs in assessing the effectiveness of different ML algorithms, presenting data in a structured format for easy comprehension.
Want AI Summaries of new PubMed Abstracts delivered to your In-box?
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!