Open ttimbers opened 2 years ago
First, I want to say congratulations to all of you! I really enjoyed your project and the general structure of your project. I know it's not easy to create such a project so kudos to all of you. Keep it up!
Please feel free to reach out to me if something is not clear. (clichyclin@gmail.com)
Again, congratulations!
This was derived from the JOSE review checklist and the ROpenSci review checklist.
1.5
This was derived from the JOSE review checklist and the ROpenSci review checklist.
2.5
list_cor
function inside the r script I have no iead what this fucntion using for with out take a clear look at the code for that. a simpole documentation will be nicemake clean
first instead of run make all
first, Since we want to remove all possible pre existed document/files, and you definitely dont want those file to influence the reproduceability. Overall it is a wonderful project I liek it, but there are some minor things need to changed.
Nice work.
This was derived from the JOSE review checklist and the ROpenSci review checklist.
Submitting authors: @anamhira47 @tonyliang19 @isabelalucas @snowwang99
Repository: https://github.com/DSCI-310/DSCI-310-Group-8
Abstract/executive summary: In this project, we will explore and predict students' exam performance about Electrical DC Machines based on their study time by using linear regression (LN) and the K-nearest neighbors (K-NN) algorithm. This result could help students gain insight into the necessary study time for specific scores as well as help instructors better understand the performance of students.
As a result of our analysis, we have found the Root mean square prediction error(RMSPE) for our LN model to be 0.281, while the RMSPE of the K-NN model is 0.257. Both types of regression have a prediction error percentage of about 40% (therefore our accuracy is about 60%), although the K-NN model is slightly better than LN model here.
This can be attributed to the fact that exam performance could be affected by other external factors such as health condition, student IQ, stress levels, learning ability and our data set may not be big enough to directly draw a relationship between just study time and exam performance.
The dataset we used was the User Knowledge Modeling Dataset provided by UCL Machine Learning Repository.
Editor: @ttimbers
Reviewer: @rpeng35 @clichyclin @harrysyz99