Modelling a desired target accurately depends on the context and nature of the target you're working with. Here's a general overview of software options and approaches based on different scenarios:

1. For Mathematical and Statistical Models:

• MATLAB: Widely used for mathematical modelling, simulations, and algorithm development. MATLAB's toolboxes and functions can be tailored to model complex systems, analyze data, and optimize solutions.
• R: A powerful statistical programming language for data analysis, statistical modelling, and visualization. R is robust in handling large datasets and complex statistical techniques.
• Python (with libraries such as NumPy, SciPy, and scikit-learn): Python, with its extensive libraries, is highly versatile for mathematical modelling, data analysis, and machine learning. Libraries like TensorFlow or PyTorch can also be used for deep learning models.

2. For Engineering and Physical Systems:

• ANSYS: Used for finite element analysis (FEA), computational fluid dynamics (CFD), and other simulations. ANSYS is ideal for modelling physical phenomena and engineering problems.
• COMSOL Multiphysics: Provides a robust platform for modelling and simulating various physical processes using multiphysics approaches. It's beneficial for complex coupled systems.

3. For Supply Chain and Operations Management:

• AnyLogic: A simulation software for modelling and analyzing complex supply chains, logistics, and operations systems. It supports agent-based, system dynamics, and discrete-event modelling.
• IBM ILOG CPLEX Optimization Studio: An optimization software for solving complex mathematical optimization problems in logistics, supply chain management, and other domains.

4. For Business and Financial Modeling:

• Microsoft Excel (with advanced plugins or VBA scripting): Excel remains a powerful tool for financial modelling, scenario analysis, and data visualization, especially with advanced plugins or scripting.
• Tableau: Useful for data visualization and creating interactive dashboards to model and analyze business trends and performance.

5. For Machine Learning and AI:

• TensorFlow: An open-source library for deep learning and machine learning models developed by Google. It's well-suited for developing neural networks and large-scale machine learning applications.
• PyTorch: Another popular open-source library for machine learning and deep learning developed by Facebook. It is known for its dynamic computation graph and ease of use in research and development.

How to Ensure Accurate Modeling:

Each software has strengths and is best suited for particular modelling tasks. Selecting the right tool depends on your project's specific needs and the nature of the target you aim to model.

Der Ohnemus

Your question has a simple answer: As per my knowledge, the only software out of the +200 MCDM in existence that can do that, is Linear Programming (LP), and in reality, it is a very simple procedure.

Think that whatever your target, you need to work with resources, that is, you need to adequate your demands, as everything in life, with your possibilities to reach the target.

Think, that if your target is to purchase a Mercedes Benz there is no way you can do that if you do not have a resource: Money to pay for it.

If a power house is delivering 400 MWh you can pretend that it delivers 500 MWh.

If you are producing product that produce contamination, above the legal limits, you cannot fabricate them

If you want to minimize costs of whatever good, you cannot pretend to have a cost less that necessary to produce the good

This means that you necessitate making comparisons between what you wish and the means or resources to reach that wish.

There is only one way to make this comparison and it is using linear inequations instead of linear equations, and only LP does this.

In addition, the LP algorithm gives you two values: The mathematical value for each criterion that corresponds to the result using the values in the corresponding criterion and from the initial matrix, and the wishes value. Then you can compare and even, for each criterion, determine in which extent your objective has been reached. For this, you need SIMUS.

Contact me if more information is needed

The most accurate software for modeling a desired target depends on the specific application. However, popular options include:

1. MATLAB (mathematical modeling and simulation)

2. Simulink (dynamic system simulation)

3. ANSYS (finite element analysis)

4. Autodesk Inventor (3D modeling and simulation)

5. OpenFOAM (computational fluid dynamics)

These software utilize algorithms, numerical methods, and machine learning to accurately model targets, incorporating factors like:

- Physics-based modeling

- Data-driven modeling

- Optimization techniques

- Uncertainty quantification

Choosing the right software depends on the target's complexity, scalability, and required accuracy.

A small addition to the LP review done by Nolberto Munier is that if there is a need for nonlinearity in a model, then KKT-conditions should be used, provided that the case has differentiability, otherwise any metaheuristic method should be used, such as "cma-es" for example.

https://www.mathworks.com/matlabcentral/fileexchange/52898-cma-es-in-matlab

Dear Steftcho

I agree with you, however Ohnemus is asking for an certain feature as is targets. Most probably MATLAB can solve the problem, I do not know, My response to Ohnemus is based on my own experience working with targets and a quantitative measure of achievment in serious problem, and in many years in OR I never saw a case solved.

Remember that it means considering simultaneously all criteria or objectives and alternatives.

In addition, we cannot do much with metaheuristics since there areseveral solutions.

Your comment about LP using only lineal data is absolutelly true, but remember that the Simplex also provides a solution to non-linearity, very restricted indeed, since it gives the Lagrange coeffcient, which of course is valid only in a point of the non-lineal objective

But of course, you also use the Khun and Tucker algotiitm in solvers that can work with no-linearity, Frontline has a solver for that

The software that most accurately models a desired target depends on the specific application and context. Here are some commonly used types:

The accuracy of the model depends on the quality of data, the appropriateness of the chosen method, and the specific characteristics of the target being modeled.

Dear Chuck

Thank you for your valuable list of software, but I do not think that you need a software to do that

Do you need a software for establishing your production target?

No, what you need is a good initial decision matrix and fix a target according to the resources you have. What you need to analyze is if you have the necessary resources to achieve what you want, but considerin g that increasing a certasin resource, may decrease another

I learnt this many years ago, when I was consulted for a farmer about what type of crop, wheat, soy and maize, was better for him to cultivate, subject of course to land availability (120 has), sowing, harvesting, available water, prices of nutrients, etc.

I solved the problem and I got the amount of each crop to grow, however, I notice something strange, because the result showed that the farmer could only cultivate 106 has.

The answer was in front of me, in the computer screen, because the reason was that the available water was not enough for the different crops in the 120 has, this could be done only in 106 has.

Thus, the farmer established the targets according to the market and his experience, looking, of course, to maximize his benefits.

But, the target for the most valuable crop, soy, couldn’t be achieved because lack of water, albeit it was enough for the two other crops. This can be understood when you think that each crop has its own rate of water consumption. For instance, maize has a water consumption much higher that the other two crops, and this is something that must be as a criterion. Maybe, in this case of mine, maize absorbed too much water and this left soy without it.

What the software does is to try to balance or satisfy each crop, but it finds that water is not enough. It is the equivalent of a ‘too short a blanket’

This is one of the problems in practically 99% of MCDM methods. They consider that money, water, land, equipment, etc. are infinite, and thus, give a solution that is maybe unrealistic.

This is an example of why you need to incorporate resources in your calculation

I used SIMUS and the software showed in tons how much was the farmer short to achieve the target, and also showed that the available land as only 106 has when the target was to use the 120 ha.

When the farmer got this information, he had several options, like decreasing the has for maize, drill a well to use aquifer water, purchase more water time if there is municipal water provided etc., i.e., he could make cost/benefit analysis