I actually spent 3 months learning ML on my own on my leisure time. I could programme simple ANN, CNN and RNN using with just Python or with Keras while understanding what is going on with the algoriths. I had no programming knowledge but I am an Engineering graduate and a Finance master. So, I have pretty good knowledge in Calculus and Linear Algebra. If you are an absolute beginner in everything ML requires, I would say it's very difficult to have a good learning outcome in just 3 months unless you are a genius. But if you want to be able to code a neural network using higher level framework like Keras, it's totally possible in 3 months. To understand the fundamentals well, you will need much more time. Right now, I am working through the deep learning textbook co-authored by Ian Goodfellow to understand the very basics of ML. I hope that will put me in a much better position to innovate in the future.
ua-cam.com/video/egTq3IYSXRM/v-deo.html Here how to learn machine learning. How to started in machine learning, reinforcement learning, imitation learning, deep neural netw See it bro I find it easy to u the
It's a little more complicated than that; in California it's very competitive. So even though I have a BS in Bioengineering: Bioinformatics, most if not all companies worth working at want at least a masters. Having someone like Siraj for Bioinformatics would be a nice thing to have while I get ready for grad school.
Take UC San Diego's course "Learning how to lean" free on Coursera. It's a pretty short course that has some very helpful techniques. Learning is not just downloading data. It's forming meaningful pathways in your brain. That's why you need to implement this stuff yourself and not just watch lectures at 2x-3x. By all means, absorb as much information as possible, just know that you need to put in much more effort than that if you want to succeed in ML. I've watched tons of lectures at 1.75 - 2.75x. It's definitely effective, but don't kid yourself. If you aren't fully grasping the materials, then try different methods of learning and different sources before you move on. Don't let the timeline force you to skip important subjects. There's a reason so many machine learning experts have masters and phds. It's not required, but it represents the level of tenacity you'll need to get intuition about ML as a whole. If that discourages you, then pick another field. If it doesn't, and it shouldn't, then GO GET IT! Take these courses, read papers, implement algorithms, watch Siraj, teach others. You're at the base of the mountain, the higher you go, the better the view gets. This curriculum isn't all it takes to get near the top. But you don't have to be at the tippy top to get a beautiful view. Thanks for the good content Siraj!
You are a great teacher man i just love when you want to start learning something and you dont know where to start and someone like you explains all of it, but your memes are outdated.
My MAN! just became subscriber! I like your teaching style but most of all your POSITIVE ATTITUDE! Yes this world belongs to those who are willing to learn! Adapt or perish :) Hello from New Zealand
I remember asking for this question and I'm so happy that you give us a deep road map so Thanks You Again you're the kind of great man that our health needs , THANKS Sir
Those who actually followed this curriculum know that doing Linear algebra alone could take upto a month easily (If done properly). Maybe Sriraj made this video intentionally as a bait to get us started learning ..
Exactly, if you want to actually understand something like linear algebra well enough to apply it to something like ML it will take you AT LEAST a month. I haven't watched the rest of the video but the same applies to calculus, I would say I'm pretty talented at math (I never had to study it at school and I learned relatively advanced stuff at a young age on my own) and it still took me 2 months for calculus 1 and 2 to get to a point where I had the confidence to actually be able to fully understand and derive something like a gradient descent for a neural network (obviously those 2 months also included all the other calculus topics, but still)
You have changed my life dude. I'm a chemical engineering student and I love this shit. the field is too young to have a dedicated degree in new Zealand and I did not want to be box myself in with software engineering when I really only want to do ml so now I feel I have a pathway without having to do that. you fucking legend I wish I could hug you.
This is the best thing you could've done to help people learn it. Alot of people get overwhelmed bx the complexity of the topic and get confussed. They start to get frustrated and quit. Very nice Model Thumbs up
Sorry, but watching math lectures at a faster speed while taking notes is DEFINITELY the way you can make sure I don't even remember the headline. I'm a slow writer and have to extremely concentrate to process thoughts someone just tells me. I have to make them my own. I can't write at the same time. I have to memorize the scentence, close my eyes, think about it, come to my conclusion and than I am ready for the next scentence. It's a slow and grinding process. I guess I'll just add a few months or years to your timeline ... Sorry if this sounded harsh or demanding or whatever ... it's just a little frustrating. I love science ... but after all these years still suck at basics like getting information into my head. Please remember there are people like me out there... I want to learn! I can learn! I'm just painfully slow.
Hi5 . I am slow to like that. I generally just speed up the lecture and pause a lot of time. Also most of my time is invested into actually thinking about the content that i just read or studied from the lecture rather than actually on the content. So maybe that should also accelarate the process. Also if you know basic Linear algebra and probability you can jump straight to ML and Python. Otherwise Math will be the grind which will be fruitfull when you start with ML and you will be breezing through everything.
Hi, i've just created a group on FaceBook #3MonthsMachineLearning_Challange_SirajRaval to apply for Siraj Raval's Challenge. Let's join this group if you're ready for this challenge. Here is the link of the group: facebook.com/groups/1991572177524785/
I think the point Siraj is trying to make is that if you want to learn extremely quickly you have to either take shortcuts (like skipping sections), or you can train yourself to absorb knowledge at a faster pace. There's ways you can trick your body to not need as much sleep, such as taking 30 minute naps for the equivalent amount of "deep sleep" you normally achieve within a night (roughly 4 hours). If you train yourself to be able to pick up concepts quickly and minimize the amount of downtime you would normally have in a class setting you can do this. I have friends in college who watch lectures in 3x speed since they are so limited in time it actually makes a difference.
This is how you learn dude. It's slow and you have to constantly repeat certain things. If you immerse yourself however, you'll eventually learn quicker because things compound. Don't stop, keep the immersion going. Good luck bro. 👍
I'm on this path. honestly 2x speed is definitely too much for me, but I understand that this video was a recommendation. I am not a math person and it's been a few years but I really want to learn machine learning. I spent 3 weeks on linear algebra. Gilbert Strang is an excellent teacher. I'm now on the first EDX course. Thank you for putting this together. It's amazing what we can learn for free.
Hi, i've just created a group on FaceBook #3MonthsMachineLearning_Challange_SirajRaval to apply for Siraj Raval's Challenge. Let's join this group if you're ready for this challenge. Here is the link of the group: facebook.com/groups/1991572177524785/
Siraj Raval guys pls make a discord server and share the link in this video that would be great or i will join today to your facebook group i am tell you because i don't use much and discord is good for group discussions
Guys just don't be dissapointed if you cannot do this in 3 months For eg. it took me about a week to get my setup done for fast.ai and get everything up and running Also like everyone else I would recommend deeplearning.ai at the start of the deep learning month.
I did not use my own hardware (920mx), I used crestle , then shifted to paperspace, and then to aws. Other users in the fast.ai community felt that a 1060 to 1080ti was enough, getting a 1080ti is recommended if you can afford it. AFAIR jeremy uses 4 1080ti SLI,
This is brilliant !!! " stop telling yourself I can't do it " part is true !! thanks for the words and such a brilliant intuitive video !! Keep up the great work !
Thanks for this encouragement, i drooped out of uni and completed a web design course 3 years ago but recently decided to start relearning all my high school math and im planning to learn advanced mathamatics so i can go back to uni next year and get my bacholers for Electrical Engineering. ive been looking for a way to test myself before hand to make sure i can handel the study load so i have been teaching myself JS, AI is something im incredibly interested in so this video has given me another goal and a great subject to elarn that will help me further my understanding. Thanks again for the brilliant content !
@@SirajRaval What about the EdX course in week 3 ... it says 18 weeks, 12 hours per week. That's 216 hours ... To do that in a week would take more than 30 hours a day. Even Elon Musk doesn't make crazy hours like that ;) You suggest to skip particular topics?
@@PACXS The hours just do not check out... he recommends 3 hours per day which is just not enough time to physically watch the videos, even at 3x speed. And on top of that he says to take notes; I can barely even comprehend most videos at 3x speed (2x is generally the highest I go), and writing down notes on top of that is just not doable. However, disregarding the suggested timeframes, the outline is pretty helpful. And at some level, everyone needs to go at their own pace e.g. some people may struggle with linear algebra but be good at calculus.
For prerequisite math: Linear Algebra, Multivariate Calculus, PCA Dimensionality Reduction, Imperial College London has just Launched a Coursera specialisation: www.coursera.org/specializations/mathematics-machine-learning
It's good but you 'll need sublimental materials if you have never been exposed to the material before hand. At least these guy tell you the more relevant stuff which is good.
This kind of stuff is the perfect format for me. I love that you took resources from all over the internet because you're legitimately picking the best ones. And I'm with you on the 2x video speed thing.
This video is an absolute gold mine. I love everything about it. I wish more people laid out material like this for other topics. I mostly program but I'm getting into investing and cooking if people know any good channels, etc. like this.
I have done something cool with recurrent neural networks, I have predicted the number of subscribers you will have, after 3.233 years, you will achieve 1 million milestone according to prediction
After all the Andrew hype I started his course first when I was a beginner, but I couldn't understand anything at first,I purchased the "Machine Learning A-Z course" on udemy and it taught me many things,then I came back to Andrew Ng course,now I understood everything. Andrew's course may not be suitable for people like me at first,but in your learning path it should be somewere there whether at first or last!
Hi Andrew Ng has already launched new courses deep learning with python. It is awesome. Small knowlege of python will help you. The new course is very fine. Check it.
Thanks so much for this video! I've just started my journey on ML and one of the difficulties I've faced is structuring my learning. Now I have a much clearer picture of my ML roadmap. Thanks and keep up the amazing work!
Thank you so much! What a great world we live in if we have people like you! Thanks for all the work you put into your videos, you're making the world a way better place!
Good job, mate. Myself, finished LA by G. Strang. Doing Multivariable Calculus @Khan Academy atm. More than half of the course made by Grant, the 3Blue1Brown guy btw. The next station is Probability, then Statistics by MIT too. How long did the probably course take?
Oh my god, I'm not interested in learning deep learning, but you have my deepest respect for compiling this information for everyone, never mind.. I'm actually getting interested..
Hi,newbie here. I'm an electronics student and my coding knowledge is very limited. I only know how to write code in C,C++. Could anybody suggest me good tutorials to get started with python? thank you. Yes,I'm very interested to learn ML and neural networks
I was just fired from my job and have nothing to do until I start with my studies this october... Siraj, you just uploaded this at the right time in my life. Let's do this!
I got an A in calc 1-3 and linear algebra and cannot vouch for 3blue1brown. His videos are fun to watch but they are not the way to really learn the material. Go for another MIT open course to learn calculus.
Will his intention is not to make you solve every integral out there or to differentiate any funcrion or whatever. His ideal is to make the math rigor that is actually very interesting and beautiful more accessible to people by presenting amazing intuitions and problems! It is a great resource to motivate yourself about math and get the well roundedness about it so that you can study a particular topic deeper. Infact, his videos on ML have lead me to want to learn it myself and build some projects as I am already a solid programmer and I know most of math behind it.
And teaching is like uploading data in your students' brain, upload speed depends on how good can you teach and how good can your students grasp! You're awesome!
Dude, this is perfect. I've basically been trying to learn ML completely backwards and just continuously getting disgruntled. After my last attempt I figured I needed to get back to the basics and really understand the knowledge behind ML instead of trying to black box everything. I didn't really know how to do this so this video is perfect timing
I used it for the first Vectors and Spaces section (minus last two) and then used www.deeplearningbook.org/contents/linear_algebra.html to figure out the other topics I had to learn and then just youtubed from there. Khan Academy gives a really good foundation.
yo bro big big fan... since I am a drop out I will try my best to learn everything in next 2 months thats because I am working on a robot at the same time (for more information please check your Instagram dm). I know how to make DNN, CNN and other stuff but I want to dive deeper and with this guide I have a ocean of resources in front of me
My friend who works in DL says this is a clickbait video. You can't learn in 3mos what it takes people currently working in the field YEARS to learn. You'll learn something but no way you can get a job.
The actual skill to apply machine learning to real life problems in an efficient way is really low currently in most companies. Getting people in who have some experience and background on the matter is pretty easy now if you go through the right channels. It will be hard to sell directly, but there are other strategies like selling your services to an established consultant company to work as an analyst that they sell forwards etc.
yeah you can learn in 5 minutes to write 10 lines to make a neural network that recognizes numbers for pictures. On the other hand you won't really understand what you will doing and the only way you will make a project is by replicating what you've already seen by changing slightly stuff [I've been there] . Siraj tries to make it look simple, but it can be really hard in the real world. To summarize you CAN do something in a few hours , you CAN'T use it for something practical in those hours
I just googled "biology" and I honestly don't know what takes people so long. I now know everything there is to animals, plants, fungi and all other living things.
I always been a guy that need to be whipped into a direction to be able to commit to something and this fits me so well damn how fortunate that this video shows on my reccomendation, thanks siraj!
Great, I might take longer than 3 months but this split up accelerated break down with ressources looks more appealing to my way of learning than anything I encountered. Thanks so much
Thank you so much Siraj!! You are literally proving us the key to unlock the door to the room of unlimited opportunity and possibility! You take the most complex cutting-edge topics and make it simple and interesting for the average programmer who wants to take advantage of the next big opportunity!
I just began my journey with Siraj 3 months plan. 12th March 2018. I will come back with bright result and some more detailed 'how to finish in 3 months' on June. As you can see, English is also hard to me like machine learning but it's not big issue. Thank you again Siraj.
Thanks Siraj for detailed guidelines. But i feel it is important that IMHO not to rush through the lectures. Solving some examples or problems definitely reinforce the maths concepts well.
Very cool suggestions man! I need to apply some of these :D
Hey! It's Jabrils! Sup man!
Cool.
Jabrils!
sometimes you wanna go... where everybody knows your na-a-ame...
Jabrils i
I actually spent 3 months learning ML on my own on my leisure time. I could programme simple ANN, CNN and RNN using with just Python or with Keras while understanding what is going on with the algoriths.
I had no programming knowledge but I am an Engineering graduate and a Finance master. So, I have pretty good knowledge in Calculus and Linear Algebra.
If you are an absolute beginner in everything ML requires, I would say it's very difficult to have a good learning outcome in just 3 months unless you are a genius. But if you want to be able to code a neural network using higher level framework like Keras, it's totally possible in 3 months. To understand the fundamentals well, you will need much more time.
Right now, I am working through the deep learning textbook co-authored by Ian Goodfellow to understand the very basics of ML. I hope that will put me in a much better position to innovate in the future.
Just the mentioning 3blue1brown deserves a like. I wish I could like more than once. Well done Siraj.
BTW, I don't think consuming a term's worth of Linear Algebra or Calculus is necessary, or can be done in a week while doing a part-time job.
I second 3B1B content is amazing!
Some quality stuff!
Indeed, 3blue1brown Essence (Linear Algebra and Calculus) are gold.
ua-cam.com/video/egTq3IYSXRM/v-deo.html
Here how to learn machine learning.
How to started in machine learning, reinforcement learning, imitation learning, deep neural netw
See it bro I find it easy to u the
I'm gonna learn from the very next month (APRIL) as I have my Board Exams now!
Thanks Siraj as always.
Ritesh Kumar Maurya same here
Glad to hear that!
Same here
Good Luck Man!
Same here !!! 10th boards ICSE.... what about you ?
I wish there was someone like you for bioinformatics here. You're god-sent to those starting in the field of machine learning.
If you find anything, could you let me know?
It's a little more complicated than that; in California it's very competitive. So even though I have a BS in Bioengineering: Bioinformatics, most if not all companies worth working at want at least a masters. Having someone like Siraj for Bioinformatics would be a nice thing to have while I get ready for grad school.
will apply ML to it in future video
Take UC San Diego's course "Learning how to lean" free on Coursera. It's a pretty short course that has some very helpful techniques. Learning is not just downloading data. It's forming meaningful pathways in your brain. That's why you need to implement this stuff yourself and not just watch lectures at 2x-3x. By all means, absorb as much information as possible, just know that you need to put in much more effort than that if you want to succeed in ML.
I've watched tons of lectures at 1.75 - 2.75x. It's definitely effective, but don't kid yourself. If you aren't fully grasping the materials, then try different methods of learning and different sources before you move on. Don't let the timeline force you to skip important subjects.
There's a reason so many machine learning experts have masters and phds. It's not required, but it represents the level of tenacity you'll need to get intuition about ML as a whole. If that discourages you, then pick another field. If it doesn't, and it shouldn't, then GO GET IT! Take these courses, read papers, implement algorithms, watch Siraj, teach others. You're at the base of the mountain, the higher you go, the better the view gets. This curriculum isn't all it takes to get near the top. But you don't have to be at the tippy top to get a beautiful view.
Thanks for the good content Siraj!
Great analogy.
Kindly Send the link for the above course.
Great 👏🏽
"the level of tenacity you'll need" - word
Tq sir it's much helpful as well as motivated..
Tq once again...
You are a great teacher man i just love when you want to start learning something and you dont know where to start and someone like you explains all of it, but your memes are outdated.
My MAN! just became subscriber! I like your teaching style but most of all your POSITIVE ATTITUDE! Yes this world belongs to those who are willing to learn! Adapt or perish :) Hello from New Zealand
I remember asking for this question and I'm so happy that you give us a deep road map so Thanks You Again you're the kind of great man that our health needs , THANKS Sir
This guy deserves an award, the fact that he spent his time to compile these resources for almost no personal gain is ridiculous. Thank you so much.
This is the best video you've ever made. I want to do it!
every single fucking time i watch his videos i become so motivated. I cant even realise who the hell is he doing .. Siraj, thank you mate.
Those who actually followed this curriculum know that doing Linear algebra alone could take upto a month easily (If done properly).
Maybe Sriraj made this video intentionally as a bait to get us started learning ..
Exactly, if you want to actually understand something like linear algebra well enough to apply it to something like ML it will take you AT LEAST a month. I haven't watched the rest of the video but the same applies to calculus, I would say I'm pretty talented at math (I never had to study it at school and I learned relatively advanced stuff at a young age on my own) and it still took me 2 months for calculus 1 and 2 to get to a point where I had the confidence to actually be able to fully understand and derive something like a gradient descent for a neural network (obviously those 2 months also included all the other calculus topics, but still)
You have changed my life dude. I'm a chemical engineering student and I love this shit. the field is too young to have a dedicated degree in new Zealand and I did not want to be box myself in with software engineering when I really only want to do ml so now I feel I have a pathway without having to do that. you fucking legend I wish I could hug you.
normal videos:
learn machine learning in 3 minutes
saraj:
learn it in 3 centuries
This is the best thing you could've done to help people learn it. Alot of people get overwhelmed bx the complexity of the topic and get confussed. They start to get frustrated and quit.
Very nice Model
Thumbs up
Sorry, but watching math lectures at a faster speed while taking notes is DEFINITELY the way you can make sure I don't even remember the headline. I'm a slow writer and have to extremely concentrate to process thoughts someone just tells me. I have to make them my own. I can't write at the same time. I have to memorize the scentence, close my eyes, think about it, come to my conclusion and than I am ready for the next scentence. It's a slow and grinding process. I guess I'll just add a few months or years to your timeline ...
Sorry if this sounded harsh or demanding or whatever ... it's just a little frustrating. I love science ... but after all these years still suck at basics like getting information into my head. Please remember there are people like me out there...
I want to learn! I can learn! I'm just painfully slow.
Hi5 . I am slow to like that. I generally just speed up the lecture and pause a lot of time. Also most of my time is invested into actually thinking about the content that i just read or studied from the lecture rather than actually on the content. So maybe that should also accelarate the process. Also if you know basic Linear algebra and probability you can jump straight to ML and Python. Otherwise Math will be the grind which will be fruitfull when you start with ML and you will be breezing through everything.
me too. It's been almost two years that i have decided to start with ML. Until now, i'm still stuck behind that
Hi, i've just created a group on FaceBook #3MonthsMachineLearning_Challange_SirajRaval to apply for Siraj Raval's Challenge. Let's join this group if you're ready for this challenge.
Here is the link of the group: facebook.com/groups/1991572177524785/
I think the point Siraj is trying to make is that if you want to learn extremely quickly you have to either take shortcuts (like skipping sections), or you can train yourself to absorb knowledge at a faster pace. There's ways you can trick your body to not need as much sleep, such as taking 30 minute naps for the equivalent amount of "deep sleep" you normally achieve within a night (roughly 4 hours). If you train yourself to be able to pick up concepts quickly and minimize the amount of downtime you would normally have in a class setting you can do this. I have friends in college who watch lectures in 3x speed since they are so limited in time it actually makes a difference.
This is how you learn dude. It's slow and you have to constantly repeat certain things. If you immerse yourself however, you'll eventually learn quicker because things compound. Don't stop, keep the immersion going. Good luck bro. 👍
You are providing a service to humanity
For accelerated learning, I am watching your video in 2x. 😆
I'm on this path. honestly 2x speed is definitely too much for me, but I understand that this video was a recommendation. I am not a math person and it's been a few years but I really want to learn machine learning. I spent 3 weeks on linear algebra. Gilbert Strang is an excellent teacher. I'm now on the first EDX course. Thank you for putting this together. It's amazing what we can learn for free.
You say 3 Months, OK LET's DO IT
Hi, i've just created a group on FaceBook #3MonthsMachineLearning_Challange_SirajRaval to apply for Siraj Raval's Challenge. Let's join this group if you're ready for this challenge.
Here is the link of the group: facebook.com/groups/1991572177524785/
few links are clipped in fb page. update them. this is great effort :)
get it
Siraj Raval guys pls make a discord server and share the link in this video that would be great or i will join today to your facebook group i am tell you because i don't use much and discord is good for group discussions
1 month later. How are doing in your learning journey? :)
I love you for this. Open and free education in quality is so scarce these days it seems, it's nice to see it still alive
Guys just don't be dissapointed if you cannot do this in 3 months
For eg. it took me about a week to get my setup done for fast.ai and get everything up and running
Also like everyone else I would recommend deeplearning.ai at the start of the deep learning month.
If you don't mind sharing, what hardware did you use for fast.ai?
I did not use my own hardware (920mx), I used crestle , then shifted to paperspace, and then to aws. Other users in the fast.ai community felt that a 1060 to 1080ti was enough, getting a 1080ti is recommended if you can afford it. AFAIR jeremy uses 4 1080ti SLI,
This is brilliant !!! " stop telling yourself I can't do it " part is true !! thanks for the words and such a brilliant intuitive video !! Keep up the great work !
I’m learning how to learn about how machines learn.
Same !!
As an ML learner, I loved your enthusiasm. Keep going.
"If you can code Gradient Descent from scratch, literally everything else becomes easier." :-)
Is it true??
@@yashwanthnandi533 I used to believe it. Now I don't anymore ... just practice coding.
Damn bro, you still working on ML. Dedication 💯💯
Thanks Siraj! Really appreciate the curriculum. The new learning revolution is the open-source learning.
I'm not trying to sound discouraging, but this stuff will take at least 6 months minimum.
Everyone learns at different speeds and different people have different amounts of free time.
Qdouble I agree
@Qdouble But you cannot do the 10 week ML nanodegree in week 3 like he says. It's just lying and slightly frustrating for him to make that claim
maybe for some. this is just what i'd create for myself. cause im impatient AF
Even 6 months is encouraging :)
Thanks for this encouragement, i drooped out of uni and completed a web design course 3 years ago but recently decided to start relearning all my high school math and im planning to learn advanced mathamatics so i can go back to uni next year and get my bacholers for Electrical Engineering. ive been looking for a way to test myself before hand to make sure i can handel the study load so i have been teaching myself JS, AI is something im incredibly interested in so this video has given me another goal and a great subject to elarn that will help me further my understanding.
Thanks again for the brilliant content !
it seems like daily 6 to 7 hours study rather 2 to 3 hours
not if you plan properly
@@SirajRaval What about the EdX course in week 3 ... it says 18 weeks, 12 hours per week. That's 216 hours ... To do that in a week would take more than 30 hours a day. Even Elon Musk doesn't make crazy hours like that ;) You suggest to skip particular topics?
@@PACXS The hours just do not check out... he recommends 3 hours per day which is just not enough time to physically watch the videos, even at 3x speed. And on top of that he says to take notes; I can barely even comprehend most videos at 3x speed (2x is generally the highest I go), and writing down notes on top of that is just not doable.
However, disregarding the suggested timeframes, the outline is pretty helpful. And at some level, everyone needs to go at their own pace e.g. some people may struggle with linear algebra but be good at calculus.
@@PACXS lmao
@@PACXS Perhaps Siraj is not great at math.
I just wanted to say thanks to helping folks learn about ML/AI/NN etc. Clear, concise and extremely helpful!
For prerequisite math: Linear Algebra, Multivariate Calculus, PCA Dimensionality Reduction, Imperial College London has just Launched a Coursera specialisation: www.coursera.org/specializations/mathematics-machine-learning
dope
It's good but you 'll need sublimental materials if you have never been exposed to the material before hand. At least these guy tell you the more relevant stuff which is good.
Thanks man
This kind of stuff is the perfect format for me. I love that you took resources from all over the internet because you're legitimately picking the best ones. And I'm with you on the 2x video speed thing.
This video is an absolute gold mine. I love everything about it. I wish more people laid out material like this for other topics. I mostly program but I'm getting into investing and cooking if people know any good channels, etc. like this.
Thank you Siraj for sharing this with us :)
one of the best motivation and learning videos i saw on youtube so far
I have done something cool with recurrent neural networks, I have predicted the number of subscribers you will have, after 3.233 years, you will achieve 1 million milestone according to prediction
ZEUS MYLIFE_2.0, it is machine learning, because I used recurrent neural networks, not that
AFAIK neural networks work very poor on extrapolation.
Deep shit!
How u got the subscriber joining rate ?
Well I'm getting a 10 mil estimate
Learning is a never ending process. It doesn't matter how old you are and where you are from. Good one
Hmmm surprised there wasn't Andrew Ng, nevertheless a cool compilation Siraj ❤
Evn i expected Adrew Ng ( coursera)
Pawan Snj May be because it's "Machine Learning" and Andrews ML course is old and uses MATLAB.
The DL course however is great.
After all the Andrew hype I started his course first when I was a beginner, but I couldn't understand anything at first,I purchased the "Machine Learning A-Z course" on udemy and it taught me many things,then I came back to Andrew Ng course,now I understood everything. Andrew's course may not be suitable for people like me at first,but in your learning path it should be somewere there whether at first or last!
Nands Pal yep!!!!! Coursera DL is the best!!!
Hi Andrew Ng has already launched new courses deep learning with python. It is awesome. Small knowlege of python will help you. The new course is very fine. Check it.
Thanks so much for this video! I've just started my journey on ML and one of the difficulties I've faced is structuring my learning. Now I have a much clearer picture of my ML roadmap. Thanks and keep up the amazing work!
Hi! Can you do one for Data Analysis/Data Science?
Finally Siraj is pacing his speech. Thank you so much bro. Now I can get you full.
This dude is crazy
Took me like 1 years to learn all those topics
have you learned all
I’m crying, thank you so much 🙏🏻 I needed this
hey Siraj! any reason why you didn't list Andrew Ng's Stanford Course? thanks
He has had a little spat with ng in the past.
@@kantonio1000 what happened??
A great demo for freshers, this video will help freshers more
Psh 3 weeks at most. Thanks for the guidance man
thats the spirit
Thank you so much! What a great world we live in if we have people like you! Thanks for all the work you put into your videos, you're making the world a way better place!
"Watch these lectures at 2x or 3x speed". "Make sure you write you own notes while watching them". How fast do you think people write notes? Jesus
Thanks, Siraj! Always good to see you in action!
2019 anyone? i've made it to Probability btw
Good job, mate.
Myself, finished LA by G. Strang.
Doing Multivariable Calculus @Khan Academy atm. More than half of the course made by Grant, the 3Blue1Brown guy btw.
The next station is Probability, then Statistics by MIT too.
How long did the probably course take?
Whoah, did you finish the roadmap? Was everything as helpful as suggested?
I'm really excited! Thank you very much for this suggestion. Now I'll study and practice machine learning not only 3 months but many more.
please make video on how to get started with github?
look for "the coding train" with daniel shiffman. He has a great starter
Oh my god, I'm not interested in learning deep learning, but you have my deepest respect for compiling this information for everyone, never mind.. I'm actually getting interested..
Siraj: You're a wizard, Prudvi!
Me: I'm a what?
This is awesome! True, anything can happen in 3 months. You make it look like everything is super easy to absorb. Your channel is great!
Hi,newbie here. I'm an electronics student and my coding knowledge is very limited. I only know how to write code in C,C++. Could anybody suggest me good tutorials to get started with python? thank you. Yes,I'm very interested to learn ML and neural networks
start with python docs
I was just fired from my job and have nothing to do until I start with my studies this october...
Siraj, you just uploaded this at the right time in my life.
Let's do this!
thank you so much
Anytime
i have udacity's 6 machine learning nanodegree programs at cheap price only $200 , mail me; biswajitxi1998@gmail.com
Please keep making amazing content Mr. Raval, you're an inspiration and incredibly generous!
I got an A in calc 1-3 and linear algebra and cannot vouch for 3blue1brown. His videos are fun to watch but they are not the way to really learn the material. Go for another MIT open course to learn calculus.
gotta do what works for you
I find it works for me he really breaks down the topic and makes it well understood
Will his intention is not to make you solve every integral out there or to differentiate any funcrion or whatever. His ideal is to make the math rigor that is actually very interesting and beautiful more accessible to people by presenting amazing intuitions and problems! It is a great resource to motivate yourself about math and get the well roundedness about it so that you can study a particular topic deeper.
Infact, his videos on ML have lead me to want to learn it myself and build some projects as I am already a solid programmer and I know most of math behind it.
I would suggest people watch his videos for intuition if they're confused about why they're doing something.
MIT OCW is not available in my country. On youtube and on site too!
So how can I learn according to you from MIT open courses?
Help me out if you can!
And teaching is like uploading data in your students' brain, upload speed depends on how good can you teach and how good can your students grasp! You're awesome!
happy holi siraj
You have saved my life ,,, I LOVE YOU MAN for this video siraj....thank you
Happy holi siraj...!
Dude, this is perfect. I've basically been trying to learn ML completely backwards and just continuously getting disgruntled. After my last attempt I figured I needed to get back to the basics and really understand the knowledge behind ML instead of trying to black box everything. I didn't really know how to do this so this video is perfect timing
why not khan academy for linear algebra
I used it for the first Vectors and Spaces section (minus last two) and then used www.deeplearningbook.org/contents/linear_algebra.html to figure out the other topics I had to learn and then just youtubed from there. Khan Academy gives a really good foundation.
It is pretty decent for a lot of these topics: they have calculus, linear algebra and statistics/probability
Siraj, you are awesome! This guide is exactly what i was looking for over last 3 days!!!! Maaany thanks!!!!
yo bro big big fan... since I am a drop out I will try my best to learn everything in next 2 months thats because I am working on a robot at the same time (for more information please check your Instagram dm). I know how to make DNN, CNN and other stuff but I want to dive deeper and with this guide I have a ocean of resources in front of me
ishant pundir please give me your whatsapp number , You sound interesting and I think we may become friends.
ishant pundir my number is 8126894553
healthy world Are you good at machine learning?
Thanks Siraj ,
After seeing ur video it feels like learning ML will be fun ,
Let's do this .
My friend who works in DL says this is a clickbait video. You can't learn in 3mos what it takes people currently working in the field YEARS to learn. You'll learn something but no way you can get a job.
And then the same recruiters cry that there are not enough engineers.
that's why he said work on 2 projects every week
You will never learn If you dont start. I dont think this video is a clickbait because every point showed by siraj are really important in ML
The actual skill to apply machine learning to real life problems in an efficient way is really low currently in most companies. Getting people in who have some experience and background on the matter is pretty easy now if you go through the right channels. It will be hard to sell directly, but there are other strategies like selling your services to an established consultant company to work as an analyst that they sell forwards etc.
You don't know how much i love you
You can literally google machine learning and learn it in a few hours
yeah you can learn in 5 minutes to write 10 lines to make a neural network that recognizes numbers for pictures. On the other hand you won't really understand what you will doing and the only way you will make a project is by replicating what you've already seen by changing slightly stuff [I've been there] . Siraj tries to make it look simple, but it can be really hard in the real world. To summarize you CAN do something in a few hours , you CAN'T use it for something practical in those hours
I just googled "biology" and I honestly don't know what takes people so long. I now know everything there is to animals, plants, fungi and all other living things.
Every time I watch one of your videos Siraj, you blow my mind !!!! And i love it !! Thanks
Is this a joke Siraj?
its the real deal
I always been a guy that need to be whipped into a direction to be able to commit to something and this fits me so well damn how fortunate that this video shows on my reccomendation, thanks siraj!
I think this channel will change my life. Thanks Siraj !
The world need more of you bro! Love every single video :)
It is the great path for to start the machine learning from very scratch, and thank you for sharing this sir.
Going to do this starting tomorrow and post some updates in progression as of a month and a half in. Thanks for putting this all out here
Great, I might take longer than 3 months but this split up accelerated break down with ressources looks more appealing to my way of learning than anything I encountered.
Thanks so much
thankyou! yours is some of the best ML content I've found, thank you for producing it. I will be starting this course asap!
You are too good Siraj...helping people all the way.. Thanks a lot!!!
Looks great Siraj! Can't wait to follow this series as it progresses!
Excellent. This is worthy 10 mins of time I spent since morning. I am taking this challenge.
I am starting it today. Will come back here to post my progress.
and?
Thank you so much Siraj!! You are literally proving us the key to unlock the door to the room of unlimited opportunity and possibility! You take the most complex cutting-edge topics and make it simple and interesting for the average programmer who wants to take advantage of the next big opportunity!
This a university student from Hong Kong. Really appreciate it! Hope to learn more!
Starting next week I am going to start doing this. This is a pretty cool curriculum suggestion. Thanks dude
This is a great direction, thanks Siraj. Going to start from today, want to make it in less than 3. I am shooting for it.
Good Work Siraj Raval. You are sharing the knowledge selflessly that's the biggest thing . Keep it up
Siraj, you are more than amazing! You can't imagine how much you motivate me... You are a source of enthusiasm and inspiration.
Thank you so much
I just began my journey with Siraj 3 months plan. 12th March 2018. I will come back with bright result and some more detailed 'how to finish in 3 months' on June. As you can see, English is also hard to me like machine learning but it's not big issue.
Thank you again Siraj.
How is it going so far?
Yo Siraj is the illest, been looking for where to start all this while. Thanks a million Karat!!!
You're amazing for giving out all of this information for free.
Thanks to you I finally found the field that I'm actually passionate about.
I like your attitude! You are honest and practical. I'm your follower now
awesome video siraj, thank you for clearing my path as well
Thanks Siraj for detailed guidelines. But i feel it is important that IMHO not to rush through the lectures. Solving some examples or problems definitely reinforce the maths concepts well.
This. This is exactly what I needed. Thank you so much, Siraj. *Subscribed!*