Dear the experts,
I have seen a lot of studies using deep learning into tomographic imaging, and improving the computational time of it.
I wonder if there is another measure for comparing computational costs.
If I am right, I found that FLOPS is widely used metric in deep learning models.
But in the iterative methods, it would not be appropriate because there can be the computations beyond the simple multiplication and summation.
If there is a good way to, I would like to ask you about it.
Thank you in advance
You could always use comparative statistics.
Have machine X run both methods to solve the same problem.
Choose one method on a single core and have its cost = 1.
You then compare all methods to this number.
As an example, your machine runs the iterative process on a single core in 1.235 hour. This is your 1 value. On two cores, the process only takes 0.7 hours. You would label this as 0.7/1.235. Keep doing this for all different processes.
Dear Mr. Park, Hyun Kyu Park
Iteration always takes a large time for computation. If you focus on computational complexity of an algorithm then it is easy to find out the computational time. Suppose you are using linear search algorithm and its complexity is O(n) and for bubble sort its complexity is O(n*n). Now apparently you can say which one take long time. For deep learning, it depends on network which you modified for your work. So please share your network and data set otherwise it is tough to compare these two.
Tank you.
- Badges
- Science topic