I took this dataset from an open dataset repository! Cool stuff, I tried a bunch of commercial grade EEG dataset during my Master and they are of varying resolution. You just have to be aware of that. What you have with the muse and emotiv is enough to make a cool machine learning project, but you will need to collect data on yourself about the state you want to detect. What project did you have in mind?

Thats right, but you have to swap the “read the research paper” for “read the documentation”.

Haha it’s that time of the year 🧔🏾‍♂️

Glad you enjoyed it🌹

Yes I live in Montreal! 😄

Depends what you mean by this: 1. If it’s linear algebra, there is no prerequisite. You can just start. 2. If it’s deep learning then you should at least know programming. But it all depends on why you are using deep learning for and what you are doing.

pre-algebra, introductory algebra, precalculus(college algebra & trigonometry), calculus 1 through 3, proof writing(alternative is discrete math)

Awesome work, keep it up! Great start to 2025! 👏👏👏

Good question, it really depends on your use case and setup. EEG is usually pretty computationally intensive to work with if you are doing lots of analysis and when you have multiple participants. What’s your use case, are you working in a lab?

@@deeplearningexplained I'm really appreciative of your quick reply! Yes, I'm working in a lab. The lab has a machine learning desktop computer but it's been used by another lab member. So I think it'll be more convenient if I have a personal laptop myself. I'm planning to classify data of fewer than 100 participants based on several experimental conditions (around 4). I still do not have a plan as to which feature to extract but I'll explore features in time, frequency, and time-frequency domains. Also, I think I'll be using Python instead of MATLAB as it seems to me that there are more accessible resources based on Python. In fact, I went through some posts and some good and up-to-date KZbin videos on machine learning laptops (e.g., kzbin.info/www/bejne/iqDaY2SupLKbgpI). One recommended laptop is Legion pro 7i. But as EEG machine learning is not their focus and I'm new to machine learning, I'm not sure if their suggestions are the best for my use case. So I think it'll be the best if I reach out to experts like you for help. Thank you, Yacine!

@@samuelsze556 Okay nice, first off do you know if your lab has access to an high performance compute cluster? If yes, you would be able to cut down the compute time of your analysis by order of magnitudes.

@@deeplearningexplained Thank you for your suggestion! We have access to a high performance facility. However, the access is restricted to full time faculty members, so I cannot use it. Do you think no laptop is suitable for my use case? Thanks again!

@@samuelsze556 A laptop would work fine, but if you can wiggle an access through your PI for the HPC that would 100X your productivity. If you have access to a HPC any laptop will do because you will use it to tweak your code and then just pull it from inside your HPC instance. If you really have to run stuff internally I would go for something with 2-4 CPU cores. Threading your code properly is one of the biggest speed up with EEG data imo. With a HPC I was able to bringdown the time it took to gather results on all our participants for a study from 1 week to 30 minutes. This was massive.

Glad it was useful! Next session is this Saturday January 4th. Happy new year btw!

Good question, what’s your end goal? Are you aiming for a job in ML?

​@@deeplearningexplained I'm actually aiming for a job as a researcher

Great question and yes that’s totally fine. However, what you should do is mark this particular problem with a red in your tracking sheet. You should come back to that problem later on to understand what exactly you are missing to solve it.

@deeplearningexplained thanks for the answer! Got it

Thanks will do, next one is on Saturday and I’ll get the answers so we can move through the material faster.

Glad you enjoyed it!

Glad you enjoyed! 🌹

Will do, I’ll improve my setup for next week session so it’s easier to follow too!

I'm answering in his place, you should try to read the book and try to solve the problems in at least the first chapter, if you're having trouble you'll need other resources, otherwise you can carry on :)

Depends on your goals, but if it is to understand the building block of ML yes!

Thanks 🌹

Yes, every Saturday I’ll stream a work session with the chapter exercises! You can also join the discord for discussion about the book!

@@deeplearningexplainedwhat is your discord channel?

Thanks! :)

it's quite similar actually but for me it's work better if I watch and also listen when I'm first getting into something

Great approach, having context is the best way to go. One thing I like the most is when the authors are presenting their research in a conference live. It’s perfect because you get the condensed summarized version of the research + you can ask them tons of questions afterward.

Yes for sure it’s here: docs.google.com/spreadsheets/d/1amCdV-uUWjGoIuJKq1I1zCaFzX--uuJaOX9fUELKj9E/htmlview

Hey thanks for the recommendation, will do 🫡! Merry Christmas and happy new year to you too!

Merci :)!

Depends how you run it, but generally AI/ML consulting is on the higher end.

depends on how good you are at marketing and product fit, imo

For a tech product yes 100%, but honestly I would always start with consulting when dealing with businesses. Much easier to go: Consulting -> product Than straight product in B2B.

💯

Great question, I’m a firm believer of the top down approach. It’s just the most efficient way to learn. However, once you are already doing that and you flagged that a lack of formal calculus is hindering your progress I would do the green, yellow and red method specifically on that topic. Doing it this way is more efficient and more motivating because you know exactly why you’re learning calculus.

Hello there, Yes I do it's over here: docs.google.com/spreadsheets/d/1amCdV-uUWjGoIuJKq1I1zCaFzX--uuJaOX9fUELKj9E/edit?usp=sharing

Awesome work! 500 exercises is already better than most ever do for a subject. I wouldn’t necessarily do them all, however I would take 1-2h to find out the one you struggled with and add them into your list of exercises. Apply the method to them and future exercise from now on, you’ll see you will become much more efficient at finding your weak points and focusing on improving those!

@@deeplearningexplained Merci beaucoup pour ta réponse, ta chaine est une mine d'or 🙏

Best of luck with your finals! 🔥🔥🔥

Thanks for mentioning, didn’t know about them!

recommend to look into, they’ve optimized process of learning to new level

They're hella expensive for me 😢

Thanks! I’m glad it was useful! 👍👍👍