## Network Analysis and Network Optimization in SAS ViyaThis course provides a set of network analysis (graph theory) and network optimization solutions using the NETWORK and OPTNETWORK procedures in SAS Viya. Real-world applications are emphasized for each algorithm introduced in this course, including using network analysis as a stand-alone unsupervised learning technique, as well as incorporating network analysis and optimization to augment supervised learning techniques to improve machine learning model performance through input/feature creation. The self-study e-learning includes: - Annotatable course notes in PDF format.
- Virtual lab time to practice.
Learn how to
- Structure networks as matrices and in the required data format (or formats) to read network data into the NETWORK and OPTNETWORK procedures.
- Define the fundamental components of network topology, including nodes, links, self-links, link weights, node weights, and directionality to understand the different ways to construct a network.
- Compute and interpret network-level measures, including network density, diameter, and average shortest path.
- Compute and interpret centrality measures, including degree centrality, eigenvector centrality, betweenness centrality, closeness centrality, and PageRank centrality.
- Compute, apply, and interpret subnetwork analyses such as connected components, shortest paths, cycles, cliques, and community detection, among others.
- Perform network querying from graph database network structures using the PATTERNMATCH statement.
- Perform network projection to transform a bipartite network into a single network with real-world applications.
- Apply network optimization algorithms such as the linear assignment problem, the traveling salesman problem, and the minimum spanning tree, among others, to solve real-world problems.
## Who should attendAnyone interested in learning to incorporate network analysis and network optimization to provide solutions and solve real-world business challenges, including data scientists, business analysts, statisticians, and other quantitative professionals. Managers, directors, and leaders with a quantitative background are also encouraged to attend to learn how network analysis and optimization can be integrated into a broader portfolio of data science and machine learning applications.
In order to complete practices with classroom software, attendees should have basic familiarity with statistics and mathematical concepts and be comfortable programming in SAS using DATA steps. Experience using macros is helpful, but not required. This course addresses SAS Viya software. Concepts in Network Analysis- Introduction.
- Network-level concepts.
- Adjacency matrices and degree centrality.
- Introduction to the NETWORK procedure.
Centrality Measures- Introduction.
- Eigenvector centrality.
- Betweenness and closeness centrality.
- Influence centrality (self-study).
- Hub and authority centrality.
- PageRank centrality.
Analysis of Subnetworks- Connected and biconnected components.
- Maximal cliques.
- Community detection.
- Paths, shortest paths, and cycles.
- Pattern matching.
Bipartite Networks- Introduction to bipartite networks.
- Network projection.
Network Optimization- Introduction.
- Linear assignment problem.
- Minimum spanning tree.
- Maximum spanning tree (self-study).
- Traveling salesman problem.
- Minimum cost network flow (self-study).
Appendix A: Network Optimization Using the OPTMODEL Procedure- Total unduplicated reach and frequency (TURF) analysis.
- Multiple traveling salesman problem (mTSP).
- Minimum cost network flow.
Appendix B: Centrality Measures Using the IML Action Set- Introduction.
- Eigenvector centrality using IML.
- Hub and authority centrality using IML.
- PageRank centrality using IML.
VYNA01 |