Graph Neural Network contextual embedding for Deep Learning on Tabular Data
Leverage (statistics)
Categorical variable
Benchmark (surveying)
DOI:
10.48550/arxiv.2303.06455
Publication Date:
2023-01-01
AUTHORS (4)
ABSTRACT
All industries are trying to leverage Artificial Intelligence (AI) based on their existing big data which is available in so called tabular form, where each record composed of a number heterogeneous continuous and categorical columns also known as features. Deep Learning (DL) has constituted major breakthrough for AI fields related human skills like natural language processing, but its applicability been more challenging. More classical Machine (ML) models tree-based ensemble ones usually perform better. This paper presents novel DL model using Graph Neural Network (GNN) specifically Interaction (IN), contextual embedding modelling interactions among Its results outperform those recently published survey with benchmark five public datasets, achieving competitive when compared boosted-tree solutions.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES ()
CITATIONS ()
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....