To anyone taking this class in the future, I highly recommend you start watching the lectures early and look for datasets. Don’t stress too hard on searching for these. You need to specify why you think their interesting (i.e. one has a lot of features, another doesn’t). Also, try not to fall behind in the lectures. That happened to me which screwed me on the midterm. I didn’t do any better on the final because of how much content there was to remember. As I said before in my previous comment, just tough it out. Get through it and don’t look back.
I took this course this fall with ML4T, and had to drop this as it was too much work, and I was not scroing great. But as I was taking ML4T in the same semester and excelling in it, i wonder how great this course would be without the tricky grading, and rude/sarcastic TA replies in forum. This class could realy use some pointer from any Prof. Joyner class.
Sadly, I really do need to agree with you. In my opinion, this class is the flagship class of the entire OMSCS program, at least for the Machine Learning specialization. I think it really needs to be improved for the OMSCS program to reach a new level of excellence.
If they are CLEARLY wrong, like they missed a whole page of your report or something, I think they will recognize it but yes, their regrade system for this class is not great.
I remember the exams being tricky so I would say get really comfortable with the main topics and understand the differences between similar topics. Also, practice is always best so if you can find practice exams or practice questions, that always helps. With ML being such a popular class sometimes people have main question games on the web.
I got an A with a 66.67%. It was by far the most frustrating and disappointing class I’ve taken and I can validate every con in this list. I’d add that it’s mandatory to attend or listen to the office hours to get an idea for what you need to do for the assignments. Overall, if you push through the frustration and put in the effort, you’ll pass, but with an ugly finish. Also, Isbell was no help, and because of him, I won’t be taking RL.
People say RL has better TAs than ML. I just finished RL course, TAs were helpful. Overall course was tough: until the last moment I thought that I would get "c". But I managed to get "b" and this course was really interesting even though very challenging. The third project is like real research: you have to find some relatively new method, try it and report the results. It's really hard and open-ended. I would like to tackle this problem (last project) one more time when I have free time. But my next course is ML, so it will not be soon
In all honesty, KBAI and ML are really not related in my opinion. You would think they would be but KBAI is more about the types of things AI can do (the HWs are about image recognition but are more hard coded) whereas ML is about actually performing ML (the HWs have you actually implement ML techniques)
I don’t know the answer to this one, I live in America and it is certainly a masters degree here. I can’t imagine why it wouldn’t be accepted in other countries but if you want to check, email Georgia tech.
Thank you for your explanation! I have 2 questions. First, I will select this course in my first semester. Is it possible to pass it? My undergraduate major was not CS, but statistics. Second, I have an official business trip during the class. In this case, may I take the final exam after or before the business trip?
it will definitely be a tough class to do as your first class. Of course it is possible to pass, just be ready to dedicate a good amount of time to the course and make sure you know at least one coding language well. As for your business trip, the final exam is usually available for a week so you will be able to take it any time during that week. Hopefully, that gives you enough time to take the exam and go on your trip. Congratulations on getting into the program!
To anyone taking this class in the future, I highly recommend you start watching the lectures early and look for datasets. Don’t stress too hard on searching for these. You need to specify why you think their interesting (i.e. one has a lot of features, another doesn’t). Also, try not to fall behind in the lectures. That happened to me which screwed me on the midterm. I didn’t do any better on the final because of how much content there was to remember. As I said before in my previous comment, just tough it out. Get through it and don’t look back.
I took this course this fall with ML4T, and had to drop this as it was too much work, and I was not scroing great. But as I was taking ML4T in the same semester and excelling in it, i wonder how great this course would be without the tricky grading, and rude/sarcastic TA replies in forum. This class could realy use some pointer from any Prof. Joyner class.
Sadly, I really do need to agree with you. In my opinion, this class is the flagship class of the entire OMSCS program, at least for the Machine Learning specialization. I think it really needs to be improved for the OMSCS program to reach a new level of excellence.
Thanks Chris, I was looking for a review :)
Thanks Chris!
so they just want you to take whatever they gave you, if you protest, even you are on the right, you will be punished for your disobedience. WTF!
If they are CLEARLY wrong, like they missed a whole page of your report or something, I think they will recognize it but yes, their regrade system for this class is not great.
how long do we have to complete a project?
what was the main difference between midterm and final?
What helped you to do better in final? any tips?
I remember the exams being tricky so I would say get really comfortable with the main topics and understand the differences between similar topics. Also, practice is always best so if you can find practice exams or practice questions, that always helps. With ML being such a popular class sometimes people have main question games on the web.
I got an A with a 66.67%. It was by far the most frustrating and disappointing class I’ve taken and I can validate every con in this list. I’d add that it’s mandatory to attend or listen to the office hours to get an idea for what you need to do for the assignments. Overall, if you push through the frustration and put in the effort, you’ll pass, but with an ugly finish. Also, Isbell was no help, and because of him, I won’t be taking RL.
People say RL has better TAs than ML. I just finished RL course, TAs were helpful. Overall course was tough: until the last moment I thought that I would get "c". But I managed to get "b" and this course was really interesting even though very challenging. The third project is like real research: you have to find some relatively new method, try it and report the results. It's really hard and open-ended. I would like to tackle this problem (last project) one more time when I have free time. But my next course is ML, so it will not be soon
I took ML4T already. I’m planning to take ML next. Do you recommend taking KBAI before ML?
In all honesty, KBAI and ML are really not related in my opinion. You would think they would be but KBAI is more about the types of things AI can do (the HWs are about image recognition but are more hard coded) whereas ML is about actually performing ML (the HWs have you actually implement ML techniques)
How did people get more than 100% on the midterm?
the exam was out of 120 so a 110/120
Is OMSCS evaluated as a Master's degree through WES for Canada Immigration ?
I don’t know the answer to this one, I live in America and it is certainly a masters degree here. I can’t imagine why it wouldn’t be accepted in other countries but if you want to check, email Georgia tech.
Thank you for your explanation!
I have 2 questions.
First, I will select this course in my first semester. Is it possible to pass it? My undergraduate major was not CS, but statistics.
Second, I have an official business trip during the class. In this case, may I take the final exam after or before the business trip?
it will definitely be a tough class to do as your first class. Of course it is possible to pass, just be ready to dedicate a good amount of time to the course and make sure you know at least one coding language well.
As for your business trip, the final exam is usually available for a week so you will be able to take it any time during that week. Hopefully, that gives you enough time to take the exam and go on your trip.
Congratulations on getting into the program!
Is OMSCS evaluated as a Master's degree through WES for Canada Immigration ?