Also go and check maths for data science in Geeks of geeks, the syllabus description is in good structure. 1. Linear Algebra & Matrix: Linear Combinations Vectors & Matrices Quantities Vectors Matrices Transpose Matrix Inverse Matrix Trace of a Matrix Determinant Matrix Dot Product Linear Mappings Functions Measurements Linear Mapping Composition Vector Spaces Formal Rules Algebraic structures Vector subspaces of a Linear Mapping Data Redundancy Linear dependence Basis and dimension Dimension of matrix spaces Fundamental Theorem of Linear Algebra Data Information Partition of linear mapping domain and codomain Data Partitioning Mappings as data The Singular Value Decomposition (SVD) 2. Probability & Statistics: Probability Continuous and Discrete Random variable Central Limit Theorem Probability distributions - Binomial, Poisson, Normal Statistics Mean, Median and Mode Standard Deviation and Variance Similarity measures - Pearson, Cosine, Spearman Hypothesis testing T-test Paired T-test p-value F-Test z-test 3. Calculus: Maxima and minima Mean value theorem Product and chain rule Taylor’s series Derivatives Gradients of Matrices Backpropagation Gradient Descent Algorithm Useful Identities for Gradient computation Higher-Order Derivatives Multivariate Taylor Series Fourier Transformations Area under the curve you should also have some discrete math knowledge and that is listed below. 4. Geometry & Graph Knowledge Deals with the angles, measurements, and proportions of ordinary objects To understand Distribution plots, Scatter plots, Boxplot (quartile, percentile) To visualize the graphs and ability to generate insights from them. Convex and Concave graphs and their properties
Great channel bro. Been thinking about creating content and your style inspired me 🙌🏾 You have a great way of maintaining good energy even through these technical subjects 💜
Great summary of the data science mathematics landscape. I have been a practicing data scientist for 6 years and have had to learn most of this stuff in a piece-meal, 'bottom-up' manner. It's great to have someone present the essential topics in a top-down 'helicopter' view - really useful to see where the different parts fit in to the overall picture. Thanks for putting this video together. There are scary numbers of data scientists in industry that just run the pre-canned statistics and machine learning packages in R and Python and who don't really understand the mathematics that underpins it all.!
I'm actually a dropout and have mostly worked jobs with absolutely zero math included, I've been learning python and sql and while I'm digesting slowly thus far, do you think there's really a pathway to learn math from scratch for people like me? I most likely will never reach your level but I just need to reach the "bare minimum" for this field, or for any kind of analytics really. Been spending much efforts to learn so I just don't wanna give up this soon. really appreciate the videos bro
Hey, it depends on your background. I would say the "bare minimum" is what you learn in the final years of school. For example, if you are in the UK, A level maths would be sufficient in my opinion. There is an always a pathway, it may just take you a bit longer. The resources I mentioned are still applicable to people like you! You may just have to work slower and look up things you are unfamiliar with along the way. And, I know it can be tough, but you have the right attitude in not giving up! It takes time, and you have to trust that it will work in the end. Take it one step at a time, learn Python and SQL like you are, then move onto the maths sections. Let me know if you need anything else :)
@@egorhowell thank you very much for reading my comment and give me a serious answer. I did rewatch your video and will try to stick to it step by step on my own progress. I'm not very hopeful since I know this field and maybe IT in general is always fast pacing and evolving very often but in my country (Vietnam) even though it's already a trendy career path right now most colleges and universities still haven't taught it yet so at least the overall saturation and competition isn't massively high. I think in your vids you explain the key points nicely and easy to understand which is very helpful, especially for people who aren't adept or fast learners like me. Keep doing the good work and I think your channel will grow exponentially!!!
I share the same sentiment - my background is in food science, so I took calculus and some linear algebra, but this was over 15 years ago! (I learned some R too) I'm keen to do this career change because I think my entrepreneurial experience will come in handy and make me a good data scientist, but not being in school for that long and now relearning math while also learning how to code for the first time feels daunting. I'm currently doing an online MSc at St. Andrews which is great because they don't require a CS background. But for many folks like me, I think we just have to keep our head up high, and take it slowly. I am realizing I'm going at it with a slower pace than my class mates, and that's ok. Thank you for these videos, it's not as anxiety-inducing as the others I've watched haha
PS. I also have the O'Reilly books, I agree that they're great! I also got some books from Amazon to help with refreshing Linear Algebra, they're the ones by Richard Han. Basics, with lots of exercises you can practice with. My textbook for my ML Algorithms course is Machine Learning Refined: Foundations, Algorithms, and Applications by Jeremy Watt, Reza Borhani, Aggelos K. Katsaggelos
Good question. Its similar, but not quite the same. For example, as an analyst you will unlikely be building ML models. So, you will using calculus a bit less. For data analytics, the main area you should focus on is probability and statistics. Hope this helps!
I have a master's in math, but looking for things I need to review. Did you really learn linear algebra in high school? That wasn't even first year of college for me.
Hey, yeah some of it but not all. In the UK we have something A-Level Further Maths and that covers things like matrices and transformations (at least when I did it!). You study this when your 16-18. I can appreciate it may vary between countries though.
@egorhowell My school had Advanced Placement (AP) Calculus, we didn't have an AP Stats class, although it exists. No linear algebra though, that's cool. The school system in North America is a lot different though.
Hi Egor, I am stuck in choosing a BSc. CS+Ai (which doesn't have any mathematic moules and is from the university that i like)(Aberystwyth Uni), and a BSc. Data Science (which has, of course, statistics and mathematical modules from Coventry Uni). Which should i rather choose? Should I choose Aberystwyth and learn mathematics and statistical from self-study?
ah sorry to hear that. Um, I would personally go with the university that I really like tbh. If you like it more, you will probably do better as well. Obviously, take my advice with a pinch of salt as I dont know all the ins and outs. Just do what you think is right, I am sure you naturally swinging to one uni anyway!
@@egorhowell Thanks, Egor❤ do u personally think the location of the university also affects summer internship and job opportunities? and what is the data scientist job market in England like esp for internation students?
Don't think location really affects whether you get the placement or not, it just depends on the travel. Like if you're university is in Scotland, but you apply for a job in London, you probably have to go in their office a couple of times a week. The market was quite tough, but is picking up at the moment from what I've seen. In terms of international students, I am not too sure on this. I don't think it really matters, but the company would require getting a work visa for it. Again, I am not an expert on this, so don't take this as professional advice!! :)
Hi I really liked your video. However, I think that the Mathametics for Machine Leanring book is extremely complex. I come from an engineering background but trying to do data science now. Are there any other books you would recxomend as a one stop shop? Are the videos a one stop shop to learn that subject ? Any answer woild he helpful.
Hey, yeah Mathematics for Machine Learning does into a lot of detail and can be overwhelming for some people which I can appreciate. I use it more like a reference text to be honest. The practical statistics for data science is excellent, and essential Math for Data Science I have also heard good things about, but haven't used it myself yet. Hope this helps!
Also go and check maths for data science in Geeks of geeks, the syllabus description is in good structure.
1. Linear Algebra & Matrix:
Linear Combinations
Vectors & Matrices
Quantities
Vectors
Matrices
Transpose Matrix
Inverse Matrix
Trace of a Matrix
Determinant Matrix
Dot Product
Linear Mappings
Functions
Measurements
Linear Mapping Composition
Vector Spaces
Formal Rules
Algebraic structures
Vector subspaces of a Linear Mapping
Data Redundancy
Linear dependence
Basis and dimension
Dimension of matrix spaces
Fundamental Theorem of Linear Algebra
Data Information
Partition of linear mapping domain and codomain
Data Partitioning
Mappings as data
The Singular Value Decomposition (SVD)
2. Probability & Statistics:
Probability
Continuous and Discrete Random variable
Central Limit Theorem
Probability distributions - Binomial, Poisson, Normal
Statistics
Mean, Median and Mode
Standard Deviation and Variance
Similarity measures - Pearson, Cosine, Spearman
Hypothesis testing
T-test
Paired T-test
p-value
F-Test
z-test
3. Calculus:
Maxima and minima
Mean value theorem
Product and chain rule
Taylor’s series
Derivatives
Gradients of Matrices
Backpropagation
Gradient Descent Algorithm
Useful Identities for Gradient computation
Higher-Order Derivatives
Multivariate Taylor Series
Fourier Transformations
Area under the curve
you should also have some discrete math knowledge and that is listed below.
4. Geometry & Graph Knowledge
Deals with the angles, measurements, and proportions of ordinary objects
To understand Distribution plots, Scatter plots, Boxplot (quartile, percentile)
To visualize the graphs and ability to generate insights from them.
Convex and Concave graphs and their properties
They have great resources as well!
Could u also please briefly explain about Mlops, generative AI and LLMS as well.@@egorhowell
Hey, yeah I might do a video on it in the future!
@@egorhowell definitely bro
if this is how the incoming competition thinks of math I have nothing to worry about
Your video series is a god sent for someone trying to break into data science, seriously, thank you 🙏🏼
Thanks Alex, really appreciate it!
Great channel bro. Been thinking about creating content and your style inspired me 🙌🏾
You have a great way of maintaining good energy even through these technical subjects 💜
I appreciate that!
Thank you for sharing this valuable information🙏
thank you!
This is great, thank you! Really clear explanations
Thank you very much!
Great summary of the data science mathematics landscape. I have been a practicing data scientist for 6 years and have had to learn most of this stuff in a piece-meal, 'bottom-up' manner. It's great to have someone present the essential topics in a top-down 'helicopter' view - really useful to see where the different parts fit in to the overall picture. Thanks for putting this video together. There are scary numbers of data scientists in industry that just run the pre-canned statistics and machine learning packages in R and Python and who don't really understand the mathematics that underpins it all.!
Thanks and I agree! I see people throw around phrases like confidence interval, statistical significant etc. without knowing what it really means.
that i was looking for, thanks :)
glad I could help!
I'm actually a dropout and have mostly worked jobs with absolutely zero math included, I've been learning python and sql and while I'm digesting slowly thus far, do you think there's really a pathway to learn math from scratch for people like me? I most likely will never reach your level but I just need to reach the "bare minimum" for this field, or for any kind of analytics really.
Been spending much efforts to learn so I just don't wanna give up this soon. really appreciate the videos bro
Hey, it depends on your background. I would say the "bare minimum" is what you learn in the final years of school. For example, if you are in the UK, A level maths would be sufficient in my opinion.
There is an always a pathway, it may just take you a bit longer. The resources I mentioned are still applicable to people like you! You may just have to work slower and look up things you are unfamiliar with along the way.
And, I know it can be tough, but you have the right attitude in not giving up! It takes time, and you have to trust that it will work in the end.
Take it one step at a time, learn Python and SQL like you are, then move onto the maths sections.
Let me know if you need anything else :)
@@egorhowell thank you very much for reading my comment and give me a serious answer.
I did rewatch your video and will try to stick to it step by step on my own progress. I'm not very hopeful since I know this field and maybe IT in general is always fast pacing and evolving very often but in my country (Vietnam) even though it's already a trendy career path right now most colleges and universities still haven't taught it yet so at least the overall saturation and competition isn't massively high.
I think in your vids you explain the key points nicely and easy to understand which is very helpful, especially for people who aren't adept or fast learners like me. Keep doing the good work and I think your channel will grow exponentially!!!
@@AliceShisori Thank you Alice, I hope all goes well. Remember to enjoy the process!
I share the same sentiment - my background is in food science, so I took calculus and some linear algebra, but this was over 15 years ago! (I learned some R too) I'm keen to do this career change because I think my entrepreneurial experience will come in handy and make me a good data scientist, but not being in school for that long and now relearning math while also learning how to code for the first time feels daunting. I'm currently doing an online MSc at St. Andrews which is great because they don't require a CS background. But for many folks like me, I think we just have to keep our head up high, and take it slowly. I am realizing I'm going at it with a slower pace than my class mates, and that's ok. Thank you for these videos, it's not as anxiety-inducing as the others I've watched haha
PS. I also have the O'Reilly books, I agree that they're great! I also got some books from Amazon to help with refreshing Linear Algebra, they're the ones by Richard Han. Basics, with lots of exercises you can practice with.
My textbook for my ML Algorithms course is Machine Learning Refined: Foundations, Algorithms, and Applications by Jeremy Watt, Reza Borhani, Aggelos K. Katsaggelos
Really great video Egor. This is super helpful!
Thank you Harry!
Thank you! Great video :)
Thank you!!
Another great video from my friend
Thank you very much :)
Do I need to learn the same math for data analytics?
Good question. Its similar, but not quite the same. For example, as an analyst you will unlikely be building ML models. So, you will using calculus a bit less.
For data analytics, the main area you should focus on is probability and statistics. Hope this helps!
@@egorhowell thanks for reply bro, any course you recommend for probability and calculus? thanks.
Hey, the one linked in the video is good. The textbook and video from freeCodeCamp :)
Excellent video, thank you:)
Glad you liked it!
Man, I appreciate 🙏
Glad you liked it!
I have a master's in math, but looking for things I need to review. Did you really learn linear algebra in high school? That wasn't even first year of college for me.
Hey, yeah some of it but not all. In the UK we have something A-Level Further Maths and that covers things like matrices and transformations (at least when I did it!). You study this when your 16-18. I can appreciate it may vary between countries though.
@egorhowell My school had Advanced Placement (AP) Calculus, we didn't have an AP Stats class, although it exists. No linear algebra though, that's cool. The school system in North America is a lot different though.
Fair enough, do you learn things like vectors, linear transformations or matrices in high school there out of curiosity?
@egorhowell I did not, there might exist some specialized school where you do learn those things.
Maybe, but every country is different. It’s not a problem though, can always study on your own!
this maths roadmap can be for ml?
Yes, I kind've use DS and ML interchangeably. (I know they are different)
Hi Egor, I am stuck in choosing a BSc. CS+Ai (which doesn't have any mathematic moules and is from the university that i like)(Aberystwyth Uni), and a BSc. Data Science (which has, of course, statistics and mathematical modules from Coventry Uni). Which should i rather choose? Should I choose Aberystwyth and learn mathematics and statistical from self-study?
disclaimer* i cant choose DS from Aberystwyth cuz i got rejected
ah sorry to hear that. Um, I would personally go with the university that I really like tbh. If you like it more, you will probably do better as well. Obviously, take my advice with a pinch of salt as I dont know all the ins and outs. Just do what you think is right, I am sure you naturally swinging to one uni anyway!
@@egorhowell Thanks, Egor❤
do u personally think the location of the university also affects summer internship and job opportunities? and what is the data scientist job market in England like esp for internation students?
Don't think location really affects whether you get the placement or not, it just depends on the travel. Like if you're university is in Scotland, but you apply for a job in London, you probably have to go in their office a couple of times a week.
The market was quite tough, but is picking up at the moment from what I've seen. In terms of international students, I am not too sure on this. I don't think it really matters, but the company would require getting a work visa for it. Again, I am not an expert on this, so don't take this as professional advice!! :)
@@egorhowell Thank you regardless, Egor. Love your videos and your vibe❤ would love to meet you someday
Hi I really liked your video. However, I think that the Mathametics for Machine Leanring book is extremely complex. I come from an engineering background but trying to do data science now. Are there any other books you would recxomend as a one stop shop? Are the videos a one stop shop to learn that subject ? Any answer woild he helpful.
Hey, yeah Mathematics for Machine Learning does into a lot of detail and can be overwhelming for some people which I can appreciate. I use it more like a reference text to be honest.
The practical statistics for data science is excellent, and essential Math for Data Science I have also heard good things about, but haven't used it myself yet.
Hope this helps!
great video
Thank you!
love from India 🇮🇳
Cheers mate!