As someone trying to learn machine learning, this gives me somewhere to start and work towards. I am so glad this video got recommended to me. Thank you so much!
Fun fact about the linear regression project. Even without scikit, you don't need gradient descent. Look up pseudo-inverses and covariance matrices if you want to make accurate predictions without all the epochs (although it might take cubic time to run).
You are right, but even if they take more iterations, for large datasets, iterative methods like gradient descent are preferred because they are more scalable and ultimately require less resources than the normal equation.
@@user-cw7yi1ew1z Basically, you use the normal equation, set the derivative of the cost function with respect to theata to zero, and solve it. This is actually what scikit-learn does. However, with millions of rows, this approach becomes computationally expensive, which is why we use methods like stochastic gradient descent.
This is just awesome! as a self learner I do get concerned about knowledge gaps..this hands-on experience will help me. Thanks for showing a path. Waiting for next videos!
Something really fun I worked on as personal project was image colorization, a great paper that I took inspiration from was Colorful Colorization by Richard Zhang.
my dream was to understand DL to a lvl that i literally built a small yet complete ML/DL framework from scratch and not just that but it can also run on gpu and cpu . then ill build all of these modules based on it . and i did just that . saying that now is actually insane af i can't believe that i did it OMG !
bro i am in 6th semester of my btech degree and my college professors are not teaching anything about AIML . i am intrest to learn these . Please tell me from where to start learning all aiml stuff in 3 months and make projects and grab a internships and then a job in this field . i will be soo thankfull to you for your suggestion
Few years ago I trained a model based on detective stories. The result was very funny and I learned a lot about input data.... Prompt: Are you a killer? Respond: That was necessarily!
@@martind9874Why? The goal of a project is to do by yourself. He gave basic idea for a project, now it's our job to pick one and search and study for it, and then implement it. It is no fun if you just copy paste a GitHub project in your repo right ? :)
May your models never overfit: That's a real blessing right there folks
As someone trying to learn machine learning, this gives me somewhere to start and work towards. I am so glad this video got recommended to me. Thank you so much!
Fun fact about the linear regression project. Even without scikit, you don't need gradient descent. Look up pseudo-inverses and covariance matrices if you want to make accurate predictions without all the epochs (although it might take cubic time to run).
But why?
You are right, but even if they take more iterations, for large datasets, iterative methods like gradient descent are preferred because they are more scalable and ultimately require less resources than the normal equation.
@@user-cw7yi1ew1z Basically, you use the normal equation, set the derivative of the cost function with respect to theata to zero, and solve it. This is actually what scikit-learn does. However, with millions of rows, this approach becomes computationally expensive, which is why we use methods like stochastic gradient descent.
(X.TX)^-1 (X.TY) lol did it some time ago
Most libraries will actually not use gradient desecent. They usually use QR-decomposition and back-sub or SVD.
Simple but useful information, even for someone at an intermediate level who needs to start building projects for their portfolio.
Great video! Going to get started right away.
I was looking this video for ages. Thanks for showing the direction ❤❤❤
This is just awesome! as a self learner I do get concerned about knowledge gaps..this hands-on experience will help me. Thanks for showing a path. Waiting for next videos!
Something really fun I worked on as personal project was image colorization, a great paper that I took inspiration from was Colorful Colorization by Richard Zhang.
Funny how you recommend projects you spit on in the intro :) Thanks for your opinion on the value on each one.
Glad I subscribed this channel
Gosh! All these projects seem to be A+ grade projects which definetly would give an edge to a portfolio. Thank you so much for this information.
Good resources and guidance to build these projects would surely help beginners
Thanks for an amazing video. Can you please also make a similar video for data analytics projects?
Super helpful. Thanks!
amazing video, it's very help full to start a project based learning on ML ⭐⭐⭐⭐⭐
Great list! Could you please suggest a few books that may help in implementing the advanced projects? Thanks.
my dream was to understand DL to a lvl that i literally built a small yet complete ML/DL framework from scratch and not just that but it can also run on gpu and cpu .
then ill build all of these modules based on it .
and i did just that .
saying that now is actually insane af
i can't believe that i did it OMG !
ur explanation is make me open minded. Now I'm ur subscriber
This is awesome!
Please provide the git links of the projects, to refer
"And you can show it to your mom".. Hell yeah!
bro i am in 6th semester of my btech degree and my college professors are not teaching anything about AIML . i am intrest to learn these . Please tell me from where to start learning all aiml stuff in 3 months and make projects and grab a internships and then a job in this field . i will be soo thankfull to you for your suggestion
How long will it take to do the first 10 projects?
2 months at max if you're willing to give your whole day everyday.
Even the ratings themselves look like data that can be fed into a model for ANOTHER ML PROJECT
Here before you are famous, remember me bro
0.54 how to build this project? I have no idea of how it is done?
just pick any dataset available and start plotting
Thanks
Bro Where are you till now, with such information!
I was looking
how to get started
not easy
it is the most difficult, to get started
my mantra : Let's go slowly, we're in a hurry
Recently, I had completed python, now what to do next ?
Building Own Linear Regression and Neural Network From Scratch,with Mathematics
I showed the fraud project to my mom. She liked it :)
Where is the code for these projects? Please share
Few years ago I trained a model based on detective stories. The result was very funny and I learned a lot about input data....
Prompt: Are you a killer?
Respond: That was necessarily!
What I'm getting from this video is that this will make our mothers proud? Am I getting this right?
SO...Great ideas...but hoe do we get started...is there a guide for each project?....please say Yes.....
Ok,thanks❤❤❤
Will be following exact pattern which u have shared here
The real boss is building a conscious AI robot.
Where can we find the link of every project?
I'm also trying to find it, but it seems that they are not gathered in one place
@@cassiolima9639 i don't see the point of the video then
@@martind9874Why? The goal of a project is to do by yourself. He gave basic idea for a project, now it's our job to pick one and search and study for it, and then implement it. It is no fun if you just copy paste a GitHub project in your repo right ? :)
After watching this video I think I am the AI GOD. But how to become a true 😂 single AI GOD
This all needs a lot of time, but Rome wasnt built in a day
hhhhhhhh jumping into cnn is hilarious
What if I don’t have a mom
Why hindi audio track is not available
There is none like God, God is not dependant on anything, anyone nor any thoughts, He does not beget nor begotten, None like Him
Your videos are the best keep it up ❤❤
I clicked because i want to be a Ai-God 🥹