Abdullah - The theory is fairly simple to explain but not always so easy to put into practice.
If we are talking about a SDOF structure, we generally want to raise stiffness, increase k, so that Fn is outside/above the speed of the forcing frequency (oftentimes running speed). This is usually preferred since it avoids being excited during startup. Adding mass will lower Fn, sometimes used in fixed-speed machines. Which to use for your particular case depends on a lot of factors and can't be answered simply with the information you provide.
To be effective, knowing where to add stiffness or add mass is important. Modifications at nodal points (locations where there is little motion) would not be effective.
You can "visualize" how the structure moves without the need of sophisticated ODS or FEA models. You can take data during operation (preferably amplitude AND phase) along the structure and plot out its motion; or a bump test when machine is not running.
Having said all this, there are a number of issues you should be aware of before you recommend a fix: Adding stiffness will usually add mass - which can be self-defeating; What is the exciting force and is it at a single frequency or does the machine have a range of speeds; Do you suspect it is a simple reed motion or something more complex?, etc.
Others on this forum can also elaborate. I would suggest you dig deeper into this so you fully understand the problem based on your comment, "....I heard somewhere there are nodes and anti-nodes..."
Regards Jim P