analyzing data statistically

I have a data that has the following variables: 1- Network: is the network # which could be 1,2,3,4 2- Algorithm is the type of the algorithm, I code is as: 1 for collaborative filtering, 2 for social algorithm, 3 for combined approach and 4 for amplified approach 3- k is the neighborhood size which takes 4 values 5,10,15 and 20 for the same network, same algorithm and same K value, I have 5 different precision and recall values. If this will help in the statistics OR I can take the average of all of these 5 values. the previous are the independent variables and I have 3 values for precision (P@2,P@5 and P@10) and 3 values for recall (R@2,R@5,R@10) which are real numbers (continuous numbers)--> these are the dependent variables my first research question is: “Which recommendation approach works the best for each network?” taking into account the neighborhood size which precision and which recall give us the highest values (highest precision, and highest recall) so it is like thinking of it as: for each network I see for each recommendation approach which give me the highest precision and highest recall. By the way: there is a trade off between the precision and recall, while the precision decreases, the recall increases my second research question is: “Which recommendation approach and which network will generate the best prediction accuracy” In other words which combination between the recommendation approach and network will produce the best precision and recall I need somebody who have very good experience in statistics to explain for me in details what statistical measures should I use, I do not need somebody to do the work. I need to understand why the selected measure is better for my data than other measures. If I can do it I will do it, if not I will create another project for the person to do it for me. attached is a sample of the data