Really loved this article, It is an inspiration for someone just getting started in Machine learning. I also visited Jack's post on linkedin the advices I think are truly helpful.
1 question if I may ask:
what should a non-tech student trying to enter into this crowded field of ML must do to stand out from the rest, I mean like what type of project I should make, what tech stack it must have?
Because I want to shift my career to this field and I am already learning about it and also currently doing an internship where i am trying to solving Black-scholes model using PINNs
Also, one way to stand out from others is to have strong supporting skills. Eg if you have a good understanding of Data Engineering fundamentals, there may be a subset of open roles where you’d might be considered more valuable than someone who has more ML expertise. This is common at smaller companies with less established data organizations so MLEs and Data Scientists are expected to contribute towards ETL pipelines.
Just to be clear, definitely make sure you have good ML fundamentals. But having other skills can help you land that first role.
Thank you for your advice, OK so "Fundamentals must be good". Noted.
one question,
How important is knowing the source code of a ML algorithm. For example if i know the mathematics and logic of say naive bayes, do i need to know the its source code so that i can create my own modified NB ? Does going that deep required for practical ML application?
Well, I’d first draw a distinction between knowing the “source code” versus writing your own implementation. I (typically) don’t know much of the source code of ML algorithms I use. But I could write them from scratch if I needed to.
For instance, I could write a Naive Bayes model from scratch, but I don’t know the source code well for the sklearn.naive_bayes package, even though I’d use that package over one I’d write myself since it’s likely to be more performant.
That said, sometimes it is important to know the source code if you’re using the framework extensively.
But if you’re new to ML, I personally would not recommend trying to understand all the source code of all the frameworks you use, as that’s way too much information to process. Also source code for most frameworks is not easily to understand since it’s typically written for performance optimization, not readability.
But others may have other views on this. So I’d encourage you to seek alternative opinions as well.
Really loved this article, It is an inspiration for someone just getting started in Machine learning. I also visited Jack's post on linkedin the advices I think are truly helpful.
1 question if I may ask:
what should a non-tech student trying to enter into this crowded field of ML must do to stand out from the rest, I mean like what type of project I should make, what tech stack it must have?
Because I want to shift my career to this field and I am already learning about it and also currently doing an internship where i am trying to solving Black-scholes model using PINNs
I have 2 videos about the kinds of projects that help you stand out.
This is beginner- https://www.youtube.com/watch?v=O187gv9rwzQ&t=4s&ab_channel=Devansh%3AMachineLearningMadeSimple
This is intermediate- https://www.youtube.com/watch?v=A5wfVTyzv9E&t=82s&ab_channel=Devansh%3AMachineLearningMadeSimple
Thanks we'll surely look into it .
+1 to Devansh’s suggestions.
Also, one way to stand out from others is to have strong supporting skills. Eg if you have a good understanding of Data Engineering fundamentals, there may be a subset of open roles where you’d might be considered more valuable than someone who has more ML expertise. This is common at smaller companies with less established data organizations so MLEs and Data Scientists are expected to contribute towards ETL pipelines.
Just to be clear, definitely make sure you have good ML fundamentals. But having other skills can help you land that first role.
Thank you for your advice, OK so "Fundamentals must be good". Noted.
one question,
How important is knowing the source code of a ML algorithm. For example if i know the mathematics and logic of say naive bayes, do i need to know the its source code so that i can create my own modified NB ? Does going that deep required for practical ML application?
Well, I’d first draw a distinction between knowing the “source code” versus writing your own implementation. I (typically) don’t know much of the source code of ML algorithms I use. But I could write them from scratch if I needed to.
For instance, I could write a Naive Bayes model from scratch, but I don’t know the source code well for the sklearn.naive_bayes package, even though I’d use that package over one I’d write myself since it’s likely to be more performant.
That said, sometimes it is important to know the source code if you’re using the framework extensively.
But if you’re new to ML, I personally would not recommend trying to understand all the source code of all the frameworks you use, as that’s way too much information to process. Also source code for most frameworks is not easily to understand since it’s typically written for performance optimization, not readability.
But others may have other views on this. So I’d encourage you to seek alternative opinions as well.
So the thing is you should be able to write the code from scratch whenever required for different use case.
So i will now focus learning the logic behind the code and try to understand them better.
Thank you for ur advice from your experience. 😀