This topic describes the cases of Optimization Solver.
Mathematical programming solving - linear programming (LP)
We have provided the following cases to explain the concept and scenarios of LP, explain how to analyze the problems that you encounter, create models, and call Optimization Solver for solving, and then use Optimization Solver results to solve the problems.
We provide the source code (Python) in the shared projects in MindOpt Studio. You can run the code in the online Notebook.
1. Concept of LP
LP problems are the simplest problems in Operations Research, and LP is the most widely used Operations Research knowledge in actual scenarios. We may have used LP methods, but we may not know what we have done.
2. Nutrition allocation: lower costs and rich nutrition
Use optimization models to set daily diet menus. The goal of nutrition allocation is to meet various nutrition requirements and reduce total costs. 12 C, C++, or Python code choices are involved.
3. Production scheduling: Optimize flow shop scheduling to minimize the makespan
A flow shop is a classic model in the scheduling field. Specify a set of machines and a batch of workpieces. The answer that you want to solve is the order of the workpieces, and the optimization goal is to minimize the makespan and reduce costs.
4. Ad traffic distribution: balanced exposure and traffic conversion
In online video throttling scenarios, make sure that the content views of each video reach the expected number and maximize the total video views. Example: ads, notifications, and the promotion content.
5. Robust linear regression - prediction
Regression is a predictive technique that establishes the relationship between the independent variable x and the associated variable y. The independent variable is a vector, and the associated variable is a scalar. Linear regression can usually be described by the linear function y = < a,x > + b, and we can estimate the values of a and b.
6. Machine learning: multi-classification problems of support vector machines (SVMs) with L1 regularization terms
Classification is one of the most fundamental tasks in the machine learning field. The purpose is to establish the mapping between the input vector x and the categorical variable y. For example, pictures are classified into categories such as animals and plants based on their content.