Abstract: Graphs and graph data science allow you to take advantage of relationships within your data that are often overlooked with traditional techniques.
Graph embeddings enable you to capture signals within networks as vectors, thus converting high-dimensional information into a lower dimension representation. This makes it easy to take advantage of topological features in ML pipelines or other analytics, such as evaluating the similarity of sparse and heterogeneous data structures.
In this session, we’ll discuss how graph embeddings build on and enrich graph data science workflows. We’ll review various embedding algorithms, as well as highlight real-world use cases where embeddings can help translate complex data patterns into tangible business value.
Bio: Katie is a Data Science Solution Architect at Neo4j. She completed her degree in Cognitive Neuroscience at Harvard University. Passionate about people and problem solving, she transitioned to focusing on helping people and businesses leverage data for impactful outcomes. As a customer-facing data scientist, she has had the opportunity to work with large and small organizations across a variety of industries. At Neo4j she helps teams up-level their data science practice with graph data science.