Handwritten Digit Recognition on MNIST dataset | Machine Learning Tutorials Using Python In Hindi
Вставка
- Опубліковано 27 сер 2024
- ►Source Code, Data & Jupyter Notebook: codewithharry.c...
►This video is a part of my Machine Learning Using Python Playlist- • Machine Learning Tutor...
►Click here to subscribe - / @codewithharry
Best Hindi Videos For Learning Programming:
►Learn Python In One Video - • Python Tutorial In Hin...
►Python Complete Course In Hindi - • Python Tutorials For A...
►C Language Complete Course In Hindi -
• C Language Tutorials I...
►JavaScript Complete Course In Hindi -
• JavaScript Tutorials I...
►Learn JavaScript in One Video - • JavaScript Tutorial
►Learn PHP In One Video - • Learn Php In One Video...
►Django Complete Course In Hindi -
• Python Django Tutorial...
►Machine Learning Using Python - • Machine Learning Tutor...
►Creating & Hosting A Website (Tech Blog) Using Python - • [Hindi] Web Developmen...
►Advanced Python Tutorials - • Intermediate/Advanced ...
►Object Oriented Programming In Python - • Object Oriented Progra...
►Python Data Science and Big Data Tutorials - • Python Data Science an...
Follow Me On Social Media
►Website (created using Flask) - www.codewithhar...
►Facebook - / codewithharry
►Instagram - / codewithharry
►Personal Facebook A/c - / geekyharis
Twitter - / haris_is_here
I have recently started a new Python course on my English channel ProgrammingWithHarry:
ua-cam.com/channels/7btqG2Ww0_2LwuQxpvo2HQ.html
where I will be analyzing each and every submitted exercise solution in depth. So make sure to solve the exercise there if you want me to analyze your solution and want me to give some suggestions to you regarding your solution.
I will make naive bayes video very very soon!
Make sure to access the ML playlist and stay tuned :)
Thanks again
I have been trying to Crack machine learning interviews.
Please create video capture, face detection, face unlock video.
Making Chatbot, NLP, NLTK, sentiment/text analysis, face detection, are needed to Crack interviews.
Can u suggest me some machine learning book?.I'm a beginner.
Sir please can u upload a video on installing keras and tensorflow both on Spyder
its your greatness that when u got wrong you showed us also to solve it using other websites like stackoverflow. keep it up mann
My name is Manish Kumar
Thank you so much sir Mai aap ka c programming ka video dekha aur mujhe itna achha laga ki Mai aapka Sara video dekhana chhahta hu
Mai jab bhi search karta hu to Sabse pahle aaphi ka video search karta hu
Thank you so much sir mai aapka jitna tarif karu kam hai. Kyo ki mujhe aapka samjhane ka concept bahot power full laga
Import numpy as np
x=[36000] change to np.array(x.iloc[36000])
thank you vai,apna hamara bohat upakar kia..
Thanks dude, this really helped
Thank you man.. I was stuck at this point from 2 days and you got answer which I required...
bahut bahut thanx vai
i was really very frustrated due to this error
you saved my life
Thanks god
First check ::
print(type(x)) -> Its is a Data Frame not a numpy array , we should first convert x to numpy array before performing any operation
Do ::
x = np.array(x)
now we can evaluate it like ::
x[0],
x[36000],
x[65487]
It's better than writing np.array(x.iloc[36000])
hope, that it helps !!!
Thank you so much for adding this comment. I was struggling while trying to plot the data in x. Your explanation helped me to visualize this!
we can also do x=x.to_numpy()
Thank you brother
You made my day.
Bro if Only I had an Award I would give it to you...Btw Thanks bro
fetch_openml has made changes to its API so if you are accessing the dataset as a pandas Dataframe you can use [mnist = fetch_openml('mnist_784' , as_frame=False)] to load it as a NumPy array.
Really helpful 🙌🙂
Or u can also use x.to_numpy()[36000]
Thanks
thanks, bro. untill 2 hours i can't solve key errorproblem ,but then put as_frame=false ,its running.
please change:
x_train[shuffle_index], y_train[shuffle_index] = x_train.iloc[shuffle_index], y_train.iloc[shuffle_index]
And i have one more doubt, why he shuffled training data, we need to shuffle entire data i.e. 70000 not 60000.
Pls reply 🥺
//this format is correct harry bhai x_train, y_train= x_train.iloc[shuffle_index], y_train.iloc[shuffle_index]
not this x_train, y_train = x_train.[shuffle_index], y_train.[shuffle_index] please check your source code and if we are converting y_train and y_test to int and if we compare with string in y_train_2 it will give error beacuse we have converted it in int so we don't need string to compare in y_train_2 we need an int so we remove double quote i have tried it that's why i am saying this and thank you harry bhai you are an inspiration for me
Can u plz upload the output
Sir,The way of your teaching is far better than our lecturers.We learned everything from sratch without getting bored.Interesting Tutorials.
Although you are teaching word by word from oreilly book but still i appreciate because i dont have time to read that book and you are covering nearly a whole chapter.
Is it? Where can i find the book? , can you specify the name of the book?
Wow Harry jus loved it...enjoyed coding wit UA channel... Grt job🔥
you can also learn fashion mnist datasets from codingcampus.xyz
Bhai aap se bdiya aaj Tak kisi ne nahi padaya specially in ML. Bahut bdiya tarike se samjhaya aapne. Keep teaching..
Finally got it!!! Also read my comments in other video please!!!
Harry Bhai,
You are such an amazing person. Your determination and decdication is really appreciable.
Providing these vedios , in such a understandable way makes become smart boy of my class.
Beginners course for python, ML stuffs , and now JavaScript vedios are really very helpfull me and fellows like me.
Thanking you up for bringing this vedio also.On text recognition.
♥️
Job lagi kahi tumhari?
@@jaanvirathore1850 Yes
Now working as a SDE -2 at Nielsen
great work sir!!! ....BIG FAN!!!!!!!!!!!!!!!!!!
#Can you please add more content on the topic like face detection project through (openCV)
#I just want to learn each and every concept from you.
Best teacher for machine learning till date!!!!!!!!!!!!!!!!!!!!!!!!!.
When he say "Agr ap ya video dakhty ho 5,6 sall bad" i looked at date of upload and it's been 4 years Now❤❤
Key Error: in [x_train, y_train = x_train[shuffle_index], y_train[shuffle_index]]
Also getting the same error if u could resolve it just tell it
did you got the solution ???
@@hitanshishah I am getting error in some_digit = x[36000] keyerror
Plz tell how you solved it
learning from you in quarantine thanks i was doing from book and not able to understand .you made it easy
Great👍! Pls cover statistics part so as to make ML concepts very clear. ☺️
Thanks to make such an informative vedio...I watched so many vedios regarding it but it is brilliant....!!!!
Bravo man...thanks alot for your lectures.
I have been trying to Crack machine learning interviews.
Please create video capture, face detection, face unlock video.
NLP, sentiment analysis are needed to Crack interviews.
Yess, I want the same stuffs
yes same here
Use tensorflow and opencv for it. First use a harrcascade to detect the face then find a face detection dataset then just put mode.predict over there.
Yes same here bro
Can we make whatsapp group
Your videos always and usually informative and good.
Dill hit Liya bhai🤩
Sir,
NLP, NLTK, Sentiment analysis, text mining, open cv, face unlock are the topics on which interview questions are based. Can you please cover these topics?
Please continue your machine learning course
Thank you sir... Aswm teaching... Understand properly
ValueError Traceback (most recent call last)
Input In [147], in ()
1 # Train a logistic regression classifier
----> 2 clf = LogisticRegression(random_state=0).fit(x_train , y_train_2)
File ~\anaconda3\lib\site-packages\sklearn\linear_model\_logistic.py:1554, in LogisticRegression.fit(self, X, y, sample_weight)
1552 classes_ = self.classes_
1553 if n_classes < 2:
-> 1554 raise ValueError(
1555 "This solver needs samples of at least 2 classes"
1556 " in the data, but the data contains only one"
1557 " class: %r"
1558 % classes_[0]
1559 )
1561 if len(self.classes_) == 2:
1562 n_classes = 1
ValueError: This solver needs samples of at least 2 classes in the data, but the data contains only one class: False
Love you Harry Bhai !!!!!!!!
Thanks for everything
great sir. good information about MINST .thanks sir🥰🥰🥰😍😘
Expected 2D array, got 1D array instead:
array=[9 7 4 ... 1 7 2].
Reshape your data either using array.reshape(-1, 1) if your data has a single feature or array.reshape(1, -1) if it contains a single sample.
isa error yaye to ye karo
model.fit(y_train.reshape(-1, 1), y_train_2)
I don't understand why accuracy is checked for training data and not test data..somebody please help
Bhaiya I appreciate ur work...thank u again..keep making stuff like this...
2ndly it's my request if u a make video series on FLUTTER too if u get time, u explain hard things in easy way...
Thank u again 😘
refer to netninja, he has amazing flutter+firebase tutorials
Any suggestion please ?
How to Create a classifier which will classify a digit always as 'not 2' ?
He is simply asking us to create a dumb classifier where all the values of y_train_2 and y_test_2 will return False . But that is not permissible while using 'clf.fit(x_train , y_train_2)'. Using it, will throw this error : ValueError: This solver needs samples of at least 2 classes in the data, but the data contains only one class: False
What we are trying to do here is trying to create a 'digit 2 classifier'. So one thing we can do is make all values of y_train_2 and y_test_2 as 'False', for the numbers 0 to 9, excluding 6 (any number), by using the below codes (This way we can avoid the Value Error by creating two classes(True/False):
y_train_2 = (y_train == 6)
y_test_2 = (y_test == 6)
So now y_train_2 and y_test_2 will return 'True' only where '6' is present and 'False' for any other number. some_digit= x[36001] contains the label as 2 at y[36001]. clf.predict(x[36001]) will return False which is correct to the extent of the code we wrote above. Notice that after cross validating you will still get the accuracy of 98 percent(subject to how the arrays were distributed inside cross validation) for all the values of False correctly predicted.
Thankyou so much for this video. It helped a lot.
Thankyou so much Harry Bhai!!!
Please make video on face reconisation on tensor flow
learn complete tensorflow deeply from codingcampus.xyz
Such a wonderful video Harry bhaii... love this tutorial bro....
i am facing key error on this line x_train, y_train= x_train[shuffle_index], y_train[shuffle_index] how i can resolve this?
same problem with me, please let me know if you are able to solve the problem
Harry bhai es playlist me kuchh video aur add kr do plz jisse hme ml project me koi problem n ho, aapko pta hoga aapne kya miss kiya, hm to abhi begginer h. Plz make some video as soon as possible 🙏🙏
Harry Bhai please continue this playlist...
bhai please make more video about ml please yar muze shirf tumhara hi technic samaj atta he
Please make a new videos on this topic so that we can learn ML better
Amazing video, har topic to the point..👏👏👏
Bahut maza aya😊
If you want to create your own MNIST like dataset quickly and easily then take a look at the InScribe app tool for iPad using the apple pencil to create digits and drawings for classification: ml-dataset-preperation-tool.web.app
Hi Harry Sir,
for this Handwritten Digit Recognition on MNIST... can we provide our own hand-written digit and test on it?
I am trying the same, but getting error as
`ValueError: Number of features of the model must match the input. Model n_features is 784 and input n_features is 8`
I have reshape the test data into 28*28.. but still it is failing with same error.
Any suggestion please.
Code i am using after having the model created.
from PIL import Image
i = Image.open('img_of_4.png')
img = np.asarray(i)
plt.imshow(img)
img.shape = (28,28)
print(img.shape)
print(clf.predict(img))
I am getting an error in last part
from sklearn.model_selection import cross_val_score
cross_val_score(clf, x_train, y_test_2, cv=3, scoring="accuracy")
It is giving me a value error
ValueError: Found input variables with inconsistent numbers of samples: [6000, 1000]
Can anyone help???
How did you fix it ?
amazing bro!! bhai asihi malware detection par ak video banao please
bhai....aap ek project tensorflow me v bana dijiye..!! btw this one is awesome.
thnax herry bhai..!!!!
make more video....
Can you please make a video about which features to use when working with our dataset
Thank bahi bhot achi series he bahi
It's amazing video Harry sir
should we use this while asked to Design a project from the MNIST dataset to identify digit classification using any of
the classification algorithms.
Harry bhai, please make a project on handwritten text recognition, this is only on character, but please make for full text..
The starter code (for loading dataset) in the latest versions
from sklearn.datasets import fetch_openml
mnist = fetch_openml('mnist_784')
mnist 784 not found show kr rha as i m using 0.24.2 sklearn version plz help
please make more videos on this playlist ,full ml pleaseeee or more projects
This one is informative, One doubt here.
19:08 why we are shuffling only 60K instead of entire data set and then split in training and test.
@@cyrilsouza6443 python doesn't have a compiler lol 😆
Do you want me to send you the email
At line - clf.fit(x_train, y_test_2)
I am getting an error - Found input variables with inconsistent numbers of samples: [60000, 10000]
Please help me solve this error!!!
same error
Sir by changing the clf also the accuracy is close to 98% what does this mean then !?
Haris bhai saare datasets ke saath 90% when (==2) and 97% when (!=2)
U r amazing , can we have some videos on intent identification
Doing a great job 😉
i am getting error while reshaping.
any help would be appreciated.
Harry sir 👍
Harry bhai.. ek insurance fraud detection ke upar ek video banao
One more ERROR
y_train = y_train.astype(np.int8)
y_test = y_test.astype(np.int8)
y_train_2 = (y_train == 2)
y_test_2 = (y_test == 2)
Bhai please ml and deep learning me vedios please bhai please bhai
Bro you are awesome
Sir kindly upload a video about handwriting detector Application using python for sentence. I am not talk about handwriting recognition. Give a complete details how we build this App? Lets take example I want to detect the twins handwriting? I want to build this App bu using GUI interface of python.
Love you bro keep this good work go on
Sir aap ML ke baad AI course bi kare
Hum logo ko bahut help hogi
Please sir
I am new Learner from Sita pur (Uttar Pradesh)
I am biggest fan of you
Please sir create a website for open a new bank
Bhai Kya gajab cheese dikha diya
I instead used svm classifier that classifies into 0-9 digits all.
Sir, in the new versions, we can't select a row of the mnist data. It's showing me error.
x = mnist['data']
x[0]
This x[0] is the error line. Please check it🙏
use x.to_numpy()[0]
harry bhai sirf simple sa code do yar jo kisi bhi folder me image jo hogi handwritten text wli usko read krske or output terminal me batasake
nice work
Thanks Harry bhai
How to install keras and tensorflow both on Spyder pe video please karo!!!!!
Sir C++ ka bhi pura tutorial video banao please
thank you so much
if we shuffled the data , what about we will get wrong lebels due to shuffling
harrybhai sign languge translator py bnao koi project ya tensor flow kerwaa dein
Thanks bro.
cannot import name 'fetch_openml' from partially initialized module 'sklearn.datasets' (most likely due to a circular import) (C:\Users\utsav\AppData\Local\Programs\Python\Python311\Lib\site-packages\sklearn\datasets\__init__.py) Can you help me to resolve this
Sir Daily News or Update ki video bano kariye app..
Jaise
Snapchat first Data Center in mumbai..
Google spend 1.1b in Netherlands data centre..
Make video plz
hi harry bai
your videos are really good .please keep up the great work
Also could you please suggest some books to keep up with the course.
it will be very useful in supplementing the videos with book to understand more
sir maine matplotlib wala x(36000) and next line error show ho raha or iska koi solution nahi mil raha plz help me?
Sir movie recommendation projects start kro sir please
Your videos are very helpful but I am unable to download necessary libraries for python and how I can make bluetooth app using python, please help
Can you slice data for independent and dependent variables
it is only applicable for 2 but not for the other images of digits for me may i know the reason
harry bhai ML pe major project pe video banao
Thank you sir.....
Great!
Thank you sir