SAS Optimization

Algebraic, Symbolic Optimization Modeling Language: Provides a flexible and intuitive way to formulate optimization models using algebraic syntax, enabling users to express complex mathematical relationships in a clear and concise manner. This facilitates transparent use of SAS functions, allowing direct integration of data processing and analysis within the optimization models.

Powerful Optimization Solvers: Offers a set of high-performance solvers tailored for different types of optimization problems. This includes solutions for linear programming, mixed integer linear programming, quadratic programming, nonlinear programming, and network optimization. The solvers are equipped with advanced algorithms and techniques to handle large-scale problems efficiently.

Network Optimization: Includes algorithms specifically designed for network optimization, which addresses problems related to flows, paths, and connectivity in networks. These algorithms enable tasks such as finding shortest paths, maximum flows, minimum cuts, and solving traveling salesman problems. It provides diagnostic tools and optimization capabilities to analyze and improve network performance.

Black-Box Optimization: Solves optimization problems involving non-smooth, discontinuous, or non-differentiable nonlinear functions. It employs hybrid parallel algorithms, combining genetic algorithms, heuristics, and pattern search methods. This capability extends the range of optimization problems that can be addressed.

Constraint Programming: Employs advanced algorithms to efficiently solve constraint satisfaction problems. It utilizes domain reduction and constraint propagation techniques to narrow down the search space. Users can choose different search strategies like look-ahead and backtracking to find one, several, or all feasible solutions. It also allows for specifying an objective function to find optimal solutions using methods like bisection search.

Distributed Computing: SAS Optimization solvers are designed to operate within the SAS Viya environment, a distributed and scalable in-memory analytics platform. This architecture enables distributed processing of analysis and data across multiple computing nodes, enhancing performance for computationally intensive optimization tasks. It also ensures fault tolerance and high availability through its distributed nature.

Cloud-Ready Deployment: SAS Optimization offers flexible data access and can be deployed in cloud environments. This facilitates on-demand access to optimization capabilities, allowing users to leverage the platform's resources and scale their solutions based on needs. Cloud deployment also simplifies infrastructure management and provides cost-effectiveness for optimization projects.

Easy Optimization Modeling: Presents optimization modeling with a straightforward algebraic language. This makes model formulation more understandable and manageable. SAS functions are directly integrated, and different optimization solvers, covering linear, mixed integer linear, quadratic, nonlinear, conic, black-box, constraint programming, and network types, are readily available. Prototyping of custom optimization algorithms is also supported.