Enhancing Benefit Adjudication Through Graph Node Embedding, Clustering, and Outlier Detection
summary
Synthetic hospitalization adjudication data with three cohorts of ~75 patients each plus a fourth noisy group (44 extra medications, 34 ICD codes) is loaded into Neo4j and processed on Azure Databricks via the Spark–Neo4j connector. Embeddings from node2vec and GraphSAGE feed density-based clustering on the Neo4j GDS library to flag outliers; the pipeline recovers nearly all of the 75 non-cohort patients and removes noisy records before model training.
AI-written summary · about
chapters (8)
- 0:00Introduction to graph techniques
- 2:13Technical overview of AI program
- 4:17Graph data model advantages
- 8:11Graph embeddings explained
- 10:40Node2Vec vs GraphSAGE
- 12:24Evaluating embedding results
- 14:09Outlier detection with clustering
- 15:55Results and future work
related talks
32mGraph-based ML for Automated Health Care Services
42mGraph Theory Takes on the Opioid Epidemic
28mRole of Knowledge Graphs in Mental Health Diagnosis and Cure
29mEnhancing Business Process Anomaly Detection With Neo4j and Graph Neural Networks
01h 41mGraph Intelligence: Enhance Reasoning and Retrieval Using Graph Analytics
27mEnhancing Financial Analytics With Event-Driven Architectures and Graph Neural Networks
41mUsing Graph-theory to empower patients
27mDr Prakash Data Science Manager, ZS Associates at Cypher 2017