I'm new with statistics and R-studio so if there's anyone among you who can help through this, they will save me.
I have a data file containing data on whether two algorithms for deep learning model successfully solved a given task, and the number of parameters in each variant of the algorithm. It is clear that increasing the number of parameters increases the chance of success.
The data file looks like this (but of course there are more data):
"success""type""parameters"
0 "algorithm1" 12720
0 "algorithm1" 10995
...
1 "algorithm2" 19458>
I need to:
- figure out whether one of the algorithms is better, controlling for the number of parameters, and explain how I determined this.
- dermine at which number of parameters does each of the algorithms have a 50% chance of succeeding and explain how I calculated this.
Thank you for your help :)