Decomposition is breaking problems into parts and try to solve little problems. So they can be very easy to settling or arranging it.
Pattern Recognition is interpreting pattern and models, then we find a similar pattern, code script or else to occur things what we want fastly. By the way we reduce our cognitive load, this is very important because of working confidence.
Abstraction is organising data logically, the process of filtering out, needless details confuses our minds in every kind of jobs. Please don't think about details that they are contribute your experiment, not like that. In time, inplace, in season learning is best for efficiency.
Alghorythm is designing and implementing alghorythm. At the beginnig of this stage we can use Pseudocode. Flowcharts are very useful for us to make the processes in alghorythm more visible.
Thanks for reading. Best wishes.