ClearVu Analytics Optimization

Optimize your products/processes

Beyond prediction for new parameter values, the key question in applications typically aims at finding an optimal process state of product design. Often, this involves multiple, conflicting objectives (e.g., cost and quality). ClearVu Analytics provides you with an efficient tool for handling these tasks, based on using models for calculating objectives, which can then be used for optimization.

To achieve this, ClearVu Analytics offers a module for multiple objective optimization. Objectives may be minimization or maximization tasks as well as attaining a target value. In addition, constraints may be used as objectives, turning the goal into satisfying an inequality, e.g., keeping cost below a given upper limit.

By means of this approach, you can define multiple, even conflicting, objectives for an optimization. In case of conflicting objectives, the result is a set of solutions arranged on so-called Pareto frontier, rather than a single solution. Again, running an optimization does not require any expert knowledge on optimization algorithms, due to the self-adaptive capabilities of our approach.

Optimization in ClearVu Analytics offers you a range of features:

• using models to formulate objectives
• maximization, minimization and goal attainment as objectives
• mixed-integer optimization
• multiple conflicting objectives can be handled through Pareto-optimization
• extended inspection of optimization results by means of graphical visualizations