I think it is the best playlist for Machine Learning and Deep Learning. Even paid courses do not provide this type of content. Thanks, sir for this type of content.
After watching a number of videos on ML, I came across this playlist and I am really enjoying how it is structured. Starting from the very basics and teaching core concepts and terminologies at the beginning. love your content.👌👌
15:55 NOTE TO ALL, fharak padta hai !!!! You are NOT advised to use scaling first, and then use train,test,split as this is a classic example of data leakage (see a couple of vids on this).
First split the data then do scaling. If you do scaling first it will give the some information (variance) of your test set into a train set. Its like you are training your whole data set and then applying it to only selected portion of your data. Which is meaning less.
16:00 First split the data then do scaling. If you do scaling first it will give the some information (variance) of your test set into a train set. Its like you are training your whole data set and then applying it to only selected portion of your data. Which is meaning less.
Starting the "100 Days of Machine Learning Challenges" journey today! It's Day 13, and I'm committed to reporting progress every day until the challenge is completed. Let's dive into the exciting world of machine learning and push our boundaries to achieve remarkable results. Stay tuned for daily updates on my learning journey and accomplishments! Day 1: 12 Day 2: 13 Day 3: 14
Best of the best playlist in entire world amazing the things which was not cleared in 3 months you just clear concepts in just 30 minutes great explanation thank you so much sir
I was also getting error like" TypeError: axis() got an unexpected keyword argument 'y_min' " Solution:- I just run command "!pip install mlxtend --upgrade --no-deps" and then after restarting google collab , rerun all lines one by one from starting . It worked for me :)
oh my god! it all makes sense now! After spending hours on finding the best source and studying from them, I never completely understood ki "chal kya raha h!". What da F*Ck is happening, up until now. The course flow is everything, 1 1 cheez now makes sense, ki actually kaise ho rha h kaam what is what. Coz trust me, we all have seen numerous 10-15 minute long minute videos on algorithms, where everything is explained in so less time, but the point is never clear. Making a note here, as to when I started learning from this series, will for sure come back and edit it with the date, I have completed it! date started: 05/11/2023
im in for completing this 100 days of machine learning goal with you...and i wont give up until i achieve my goal at any cost..thanks for making such a good content im really loving and and i cant thanks you less for uploding and being so dedicated for proving such a fine content...thanks sir and i promise to complete my goal at any cost..
if you have progressed further in this playlist can you tell me whether vids till 50ish all about data preprocessing and can i skip those if i already have some idea about those ?
isn't it a support vector machine algorithm you yourself has said earlier that regression models are used to obtain numerical output rather than categorical output?
Sir I have a Question, at 22:42 Training Model accuracy is 90 percent, Model is predicting 2nd student wrongly, In my opinion we should look at the dataset and correct the input values, by doing this our model accuracy rate will be above 90%? Please answer my question even If I am wrong. Thank you Sir.
sir i'm not getting point on plotting in last step although i'm getting the graph TypeError: axis() got an unexpected keyword argument 'y_min' please help
Hi Sir, great video..just one thing I wanted to point out is that scaling always happens after the train test split right? because the test data is supposed to be hidden-and if we scale on that along with train data there will be some data leakage issues..which is why we do fit transform on x train and only transform on x test, pls correct me if im wrong!
Yes, you are correct. In machine learning, scaling or normalization is usually applied to the training data before the model is trained. This is done to ensure that the model is trained on a standardized range of input features. The scaling process involves calculating the mean and standard deviation of the training data, and then applying a scaling transformation to the data based on these statistics. If scaling is applied to the entire dataset before the train-test split, then information from the test set can potentially leak into the training set, resulting in overfitting and inaccurate model performance. Therefore, the recommended approach is to first split the data into training and testing sets, and then apply scaling only to the training set. After scaling the training set, the same scaling parameters (mean and standard deviation) can be used to transform the test set. This ensures that the test set remains unseen and unbiased, and the model's performance can be accurately evaluated on new, unseen data.
why i am getting y_pred and y_test completely same and during this clf.fit(x_train,y_train) i am not getting the same output like yours ...its just showing logisticRegression without a full info as yours
Hey I tried using the mlxtend library, but getting error module not found. I have already installed the library and also checked the path. Could you please help with this?
Thank you so much but yaar sach me maza nahi aya during model training, i was thinking about the model training it will be a huge graphical interface........
I am getting 100% accuracy for this dataset. But when i am plotting the decision regions, i can see some points in the incorrect region. How is this possible??🤨
First time, I have created a Machine Learning project! But my bad! The accuracy score is just 50%🥲🥲! By the way, thanks for making this helpful and immersive content in a cost of 0 rupees!
sir after this promt :clf.fit(X_train,y_train) This error is coming , plz solve or tell C:\Users\Ravi Shankar\anaconda3\lib\site-packages\sklearn\utils\validation.py:993: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel(). y = column_or_1d(y, warn=True)
Bhai ur super u have deep knowledge. I was using iloc[:,2:3] with this everything was right but same problem why bro? Can u explain pl. N after using ur soln I didn't use y-train.values n I got plot how bro.
@CampusX sir ye problem or data ka y label class hai like either 0 or 1 tow hm is me regression ka model ku use kr rhe hai classification ka use ku nhi kr rhe?
Sir why is it that whenever i run my code from top it changes it accuracy and shows different behaviour with same dataset and same splitting on same algorithm ? Is is because of size of dataset or is there any other reason behind it.
I think the main reason is your model training, every time you run your whole notebook your model try to optimized itself and try to minimize the error for e.g. In Gradient descent It iterates for no. of times and every time it try to minimize the slope, so every time you run, it maybe get better optimization and less error but there should not be much difference. If the accuracy have much difference every time you run your notebook, there could be something wrong in the code. Hope it will help.
I think it is the best playlist for Machine Learning and Deep Learning. Even paid courses do not provide this type of content. Thanks, sir for this type of content.
The confidence this lecture gives to a newbie in Machine Learning is unimaginable.
Thanks a bunch.
I think you are the best machine learning youtuber on whole UA-cam.
Thank you for sharing this valuable content. ❤️😘
After watching a number of videos on ML, I came across this playlist and I am really enjoying how it is structured. Starting from the very basics and teaching core concepts and terminologies at the beginning. love your content.👌👌
15:55 NOTE TO ALL, fharak padta hai !!!! You are NOT advised to use scaling first, and then use train,test,split as this is a classic example of data leakage (see a couple of vids on this).
Means we need to do train test split first then scaling right?
Had 1 question if we had done the same things same data can our acuuracy_score varry ?
First split the data then do scaling.
If you do scaling first it will give the some information (variance) of your test set into a train set.
Its like you are training your whole data set and then applying it to only selected portion of your data. Which is meaning less.
Exactly, I commented and then I saw this comment. It was really weird that he did not mention this
16:00
First split the data then do scaling.
If you do scaling first it will give the some information (variance) of your test set into a train set.
Its like you are training your whole data set and then applying it to only selected portion of your data. Which is meaning less.
The only teacher who gets the need and value of bird's eye view intros!!
Starting the "100 Days of Machine Learning Challenges" journey today! It's Day 13, and I'm committed to reporting progress every day until the challenge is completed. Let's dive into the exciting world of machine learning and push our boundaries to achieve remarkable results. Stay tuned for daily updates on my learning journey and accomplishments!
Day 1: 12
Day 2: 13
Day 3: 14
daily, 3 din me khatam
After leaving the hope to learn ML conceptually, this channel is blessing for. Keep Uploading videos.
Best of the best playlist in entire world amazing the things which was not cleared in 3 months you just clear concepts in just 30 minutes great explanation thank you so much sir
May you be blessed with happiness in your life for generously sharing your valuable time and content, conveyed with love from Pakistan.
3 year se learn kerraha hoon but koi samajh ni ayi you are the best teacher of ML,DL,NLP and much more 🥰🥰😍😍💗💗💓💓💖💖💗💗💝💝❣❣💟💟
Best ML Playlist in youtube especially than any indian AI Based Content creation.
So how I landed on this video, for me its amazing experience as a newbie in ML
Best strategy to show end-to-end flow before Learning.
Sir your teaching is very great , loved this playlist 😊
You are a gem. Love from Pakistan. May God Bless you Sir.
I was also getting error like" TypeError: axis() got an unexpected keyword argument 'y_min' "
Solution:- I just run command "!pip install mlxtend --upgrade --no-deps" and then after restarting google collab , rerun all lines one by one from starting . It worked for me :)
Everything is clear to me after watching your video ❤️
sir your way of teaching is best than any other ML youtuber, Thanks alot for sharing knowledge... ✌
You are amazing .... Every ML Students Should watch your video its Outstanding
Brother how to download the csv file used here?
@@nishantpatial2611 You can easily create that csv file just copy and paste in notepad file and save it with csv externsion
wah vai wah bhaut khub
This content is too good to be real. I hope im not dreaming rn
Hello Sir..Thank u soo much, very good explanation
m big fan of urs........really helpful ur vdoz are ..... great Job!!..........10x.............. :)
You are doing a great work,Thank you 😊
Great explanation any beginner could understand it ❤
Very nice content, well understanding concepts..
Thank You Sir❤
He is Legend, big Salute !!
22:01 sir made the point, I also used to imagin model training mein naa janne kya hi hota hoga, LOL
Fantastic explanation 👌👌
thank you sir it gives me interest to move forwwarn in my journey
Fantastic explaining. Could you please teach us ML in particle physics? How to start? How to get particle data online for research? Thank you
PLACEMENT HO JAYGA.. THANKS A LOT SIR
Thank you sir
Thankyou sir
oh my god! it all makes sense now! After spending hours on finding the best source and studying from them, I never completely understood ki "chal kya raha h!". What da F*Ck is happening, up until now. The course flow is everything, 1 1 cheez now makes sense, ki actually kaise ho rha h kaam what is what.
Coz trust me, we all have seen numerous 10-15 minute long minute videos on algorithms, where everything is explained in so less time, but the point is never clear.
Making a note here, as to when I started learning from this series, will for sure come back and edit it with the date, I have completed it!
date started: 05/11/2023
Have u completed?
have u completed
you are great !!!!
im in for completing this 100 days of machine learning goal with you...and i wont give up until i achieve my goal at any cost..thanks for making such a good content im really loving and and i cant thanks you less for uploding and being so dedicated for proving such a fine content...thanks sir and i promise to complete my goal at any cost..
if you have progressed further in this playlist can you tell me whether vids till 50ish all about data preprocessing and can i skip those if i already have some idea about those ?
I guess you guys should follow the recent videos of him, the data science playlist
Nice teacher
amazing explanation thankyou
Your content is great..keep up the good work
Wow! Very much helpful. Thanks
A Standard Scaler is generally performed if we have negative values in our dataset, I think we should use a min-max scaler here.
Awesome 🔥
so good video love from nepal
So easy explanation
great😇 work
legend🙇🙌
Thank you, I have question? after doing this practice ? will you teach us Facemask detection model?
Thankyou sir❤❤❤❤
Sab samajh me aagya
JAI SHRI RAAM
Very clear to understand.
isn't it a support vector machine algorithm you yourself has said earlier that regression models are used to obtain numerical output rather than categorical output?
Thank you Very Much Sir!!!
Just a question? I thought we do scaling after train-test-split to avoid data leakage? is it not the case?
Thank you sir for teaching us world class machine learning with our own language at zero cost. 🫡🙏🏻🙏🏻🙏🏻
JP Here,
Thank You Sir :)
finished watching
Sir I have a Question, at 22:42 Training Model accuracy is 90 percent, Model is predicting 2nd student wrongly, In my opinion we should look at the dataset and correct the input values, by doing this our model accuracy rate will be above 90%? Please answer my question even If I am wrong. Thank you Sir.
Date 3/11/24..
Started this today
Why we use fit_transform() on training data but transform() on the test data?
sir i'm not getting point on plotting in last step although i'm getting the graph
TypeError: axis() got an unexpected keyword argument 'y_min' please help
Thanks a lot sir❣️
Hi Sir, great video..just one thing I wanted to point out is that scaling always happens after the train test split right? because the test data is supposed to be hidden-and if we scale on that along with train data there will be some data leakage issues..which is why we do fit transform on x train and only transform on x test, pls correct me if im wrong!
Yes, you are correct. In machine learning, scaling or normalization is usually applied to the training data before the model is trained. This is done to ensure that the model is trained on a standardized range of input features.
The scaling process involves calculating the mean and standard deviation of the training data, and then applying a scaling transformation to the data based on these statistics. If scaling is applied to the entire dataset before the train-test split, then information from the test set can potentially leak into the training set, resulting in overfitting and inaccurate model performance.
Therefore, the recommended approach is to first split the data into training and testing sets, and then apply scaling only to the training set. After scaling the training set, the same scaling parameters (mean and standard deviation) can be used to transform the test set. This ensures that the test set remains unseen and unbiased, and the model's performance can be accurately evaluated on new, unseen data.
why i am getting y_pred and y_test completely same and during this clf.fit(x_train,y_train) i am not getting the same output like yours ...its just showing logisticRegression without a full info as yours
I have faced the problem, when I create the plot decision region.
thanks
Hey I tried using the mlxtend library, but getting error module not found. I have already installed the library and also checked the path. Could you please help with this?
Sir ,
Thanks for sharing such valuable information with us on youtube ,a sincere request to u is that will u pl make video on Git
and Github.
GIT and GitHub Video Link by CampusX: ua-cam.com/play/PLKnIA16_RmvbzH779EIJfNEmqgNEjnnKL.html
on scaling , every value is not between -1 to 1 then why sir is saying that every value is b/w -1 to 1 ??????
X_test = scaler.transform(X_test)
getting error on this line of code ??
please help
Sir where can I find your playlist on deployment of models in aws, heroku or gcp?..... Thank you sir🙏🙏
Haven't created it yet. Will create soon
Can you please show how you created website please
Sir ap me model deployment sa related videos ka bataya Hua ha but wo is Playlist ma available nai ha?
how is that xtrain value in range of -1 to 1, can somebody explain it to me ?
each time i am refreshing cells the accurancy is changing like 7,8,9 but sometime it giving 1 or 2 also
is it happening because of small data?
⁉⁉23:00 Why accuracy score is 0.9 even if 2 out of 10 values are wrong????
Which algo you used.
just found out my coaching teacher been following you for a long time..damn
21:38 made me laugh :)
Had 1 question if we had done the same things same data can our acuuracy_score varry ?
Thanks, sir
Why is there a need to transform the data
Why my accuracy is coming to be 1.0 even I have written the same code
Thank you so much but yaar sach me maza nahi aya during model training, i was thinking about the model training it will be a huge graphical interface........
great
But can u share your dataset link with us?
I am getting 100% accuracy for this dataset. But when i am plotting the decision regions, i can see some points in the incorrect region. How is this possible??🤨
Sir, why we didn't perform the feature scaling on y_train and y_test data ?,,,,,why only on X_train and X_test data
You can do but there was no need. It is already 0 and 1
First time, I have created a Machine Learning project! But my bad! The accuracy score is just 50%🥲🥲! By the way, thanks for making this helpful and immersive content in a cost of 0 rupees!
sir after this promt :clf.fit(X_train,y_train)
This error is coming , plz solve or tell
C:\Users\Ravi Shankar\anaconda3\lib\site-packages\sklearn\utils\validation.py:993: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().
y = column_or_1d(y, warn=True)
bhai mere main bhi aa rha hai ye error solve hua kya?
you must have written "y = df.iloc[:, -1:]", instead change it to "y = df.iloc[:,-1]".
Bhai ur super u have deep knowledge. I was using iloc[:,2:3] with this everything was right but same problem why bro? Can u explain pl. N after using ur soln I didn't use y-train.values n I got plot how bro.
Course Started : ML
Lecture-01: 14/08/2024
Lecture-02: 14/08/2024
Lecture-03: 14/08/2024
Lecture-04: 14/08/2024
Lecture-05: 14/08/2024
Lecture-06: 15/08/2024
Lecture-07: 15/08/2024
Lecture-08: 15/08/2024
Lecture-09: 15/08/2024
Lecture-10: 15/08/2024
Lecture-11: 16/08/2024
Lecture-12: 16/08/2024
Lecture-13: 17/08/2024
majaaaaa aaai
Where is the dataset?
@CampusX sir ye problem or data ka y label class hai like either 0 or 1 tow hm is me regression ka model ku use kr rhe hai classification ka use ku nhi kr rhe?
Bro log reg hi classification me lagta hai
Sir why is it that whenever i run my code from top it changes it accuracy and shows different behaviour with same dataset and same splitting on same algorithm ? Is is because of size of dataset or is there any other reason behind it.
I think the main reason is your model training, every time you run your whole notebook your model try to optimized itself and try to minimize the error for e.g. In Gradient descent It iterates for no. of times and every time it try to minimize the slope, so every time you run, it maybe get better optimization and less error but there should not be much difference. If the accuracy have much difference every time you run your notebook, there could be something wrong in the code.
Hope it will help.
How can one download the dataset..
Sir can you show backend of that website
can you plz send the code of the streamlit app you created