Hello! If one just finished from high school and wants to get started building machine learning skills during the summer before going on to start freshman year in Computer Science, is it to ambitious to start first with the Intro to Machine Learning course? Or should I start with taking the Mathematics for Machine Learning course and Introduction to Statistics course on Coursera respectively. Also, I have basic python knowledge; however, I don’t know how to use Pandas, NumPy and Matplotlib. I am on week one in the ML course and the second lab has python code that uses these libraries. Would I be fine by just learning them from YouTube and then continuing with the ML course? Resources I found:
Personally I would skip the Math for Machine Learning course. If you know about probabilities for rolling dice, and understand what a mean and a standard deviation are, you don’t need a statistics course.
Go can right to the Machine Learning Specialization. It’s an introduction, and you can learn what you need along the way.
@Ufedo I mean… not to disagree with TMosh, but I double checked on the Imperial College course (I have not taken it but I’ve been a mentor for them for two years now).
What they teach is hardly just probability-- but, all at once, you might find it a bit overwhelming.
Everything they teach though, down the line, you will find useful.
So I say ‘give it a shot’, but don’t feel discouraged if you don’t understand it all the first time either…
@Nevermnd Thank you for your thoughts. Please, by “useful down the line”, do you mean for my Comp Sci degree or this Intro to ML course by Andrew Ng or for my ML journey as a whole especially after completion of the intro course? I know I emphasize completion of the course but I want to be able to practice concepts learnt using Kaggle and build projects after and during the course. So, would this Math for Machine Learning be a prerequisite for that?
S’okay/cool beans. They are rather clever though so I can’t discount them. And you are perfectly correct – @Ufedo doesn’t need to know all of this at once to start, but he will know what he is getting into."
@Nevermnd Okay, so if I don’t need all “these” knowledge to get started with ML using this course, do I need it to successfully finish, practice knowledge using Kaggle and build projects? Please by “she will know what she is getting into”, how do you mean?
@Ufedo so TMosh is correct-- There are all sorts of libraries/languages (i.e. R Julia) out there these days such that you don’t strictly need the math background to do ML.
I guess the question is more your end goal to be a practioner of ML (i.e. someone that uses such tools to produce meaningful analytic results for clients)-- or a researcher, developing new methods and ways of thinking about these problems ?
Neither of these paths is better than the other, it just depends on your personal interest.
I mean, for myself, I always tend to be curious about/like taking apart stuff, and at some point in your journey you might find yourself curious too-- In this case the math skills will be very helpful.
As a direct example there are tons of libraries that will let you run kNN (K-Nearest Neighbors), but if you really want to know how they are actually doing this calculation, from scratch, I highly recommend the textbook:
EDx also has a direct course from MIT covering this very text.
So, strictly, all that math is not needed to start-- But later in your career I think will be very helpful (as you mentioned you are just in high school).
Thus by ‘what are you getting into’ I meant just knowing the important areas of study (with regards to the maths) and recognizing it might take a few years to get there.
Depending on your long term goal. If you want to get the “fast” overview of ML, step directly into the Machine Learning course. You’ll learn all you need, including the necessary math, on the way. If you get into problems, ask the course administrators.
However, if you want to go deeper in the basics of mathematics, and want to later to build own methods, take the other course first.
In booth cases, for better understanding of Python, ask GPT4 if you have access to it. So you’ll learn Python on the way.
I am learning as I’m learning from coursera is the machine learning specialization, after that I will go to the Kaggle to find a project relating to what I learned from the specialization. Then I want to find a job relating to that, but we need to have a solid knowledge like regression model, classification, retrieval. Math and ML idea are important
I am an incoming Comp Sci freshman planning to start internship applications this fall and am currently building Python skills through a course, 100 Days of Code by Angela Yu on Udemy, and ML skills through Andrew Ng’s specialisation on Coursera. Since tech interviews often involve Data Structures and Algorithms (DSA), can I effectively learn DSA alongside Python and Machine Learning (ML)?
I have basic Python knowledge (up to classes and objects) and familiarity with NumPy, SciPy, and Matplotlib.
Is it feasible to learn DSA while building Python and ML skills concurrently?
Which DSA course (Google’s Udacity or Zero to Mastery) would you recommend, or are there alternative resources? Google’s Udacity course is a bit pricy; with a 40% discount, it costs $150 per month to get access and the course lasts for three months.