Interview Prep Day 2- Linear Regression Interview Question-The Most Important Algorithm In ML & DS🔥🔥
In this video we will be understanding the important interview questions that are usually asked regarding Linear Regression
It is must for every data science aspirants
github: github.com/krishnaik06/Interv...
Data Science Interview Playlist: • Data Science interview...
Please join as a member in my channel to get additional benefits like materials in Data Science, live streaming for Members and many more
/ @krishnaik06
Please do subscribe my other channel for AI Tech news
/ @krishnaikhindi
Connect with me here:
Twitter: / krishnaik06
Facebook: / krishnaik06
instagram: / krishnaik06
#datascienceprojects
#machinelearninginterviews
#deeplearninginterview
#datascienceinterviews
#computervision
#nlp
Пікірлер: 80
I don't understand why these dislikes for some one who is making efforts to teach us, dedicating their time for us, and on that for FREE.... What is their to dislike in Krish's videos...it pains to see we got useless creatures disliking..
@AAND8805
3 жыл бұрын
Demotivated statisticians who think they will be replaced by Data Scientists in the near future 😁
@gssci
2 жыл бұрын
I appreciate Krish’s content too, but his videos are supported by ads and paid subscriptions, it’s not like he’s doing it just out of the kindness of his heart. Also, dislikes are not always all that bad, they may help improve the content and grow the channel even more. (A dislike or a negative comment is still an interaction that help promote the video in the suggestion algorithm)
This series will be the best amongst all. I am pretty sure that this series will accelerate us to land in field of ML & DL. Thanks Krish for all these efforts
I have no words to thank you. Your videos have helped me a great deal in understanding Data Science. I will be able to switch my career finally. I owe you a big THANKS :)
Hi sir I made a video on KZread channels for Data science. I'm happy to say that your channel also included in video And your explanation is so clear. Thank you
Thank you Krish, you are awesome. I am following most of your videos.
great sir you are my mentor i follow all your videos.
Doing great job sir!! Keep going, this will help a lot.
This is amazing content! Thank you Krish
The normality assumption for linear regression applies to the errors, not the outcome variable per se (and most certainly not to the explanatory variables). The usual statement is that the errors are i.i.d. (i.e., independently and identically distributed) as Normal with a mean of 0 and some variance.
@DeepROde
3 жыл бұрын
Correct
amazing Krish Your true motivation with learning for journey in Data science :)
Your video is very good sir, but here I want to add on more regression multilevel regression and selection of feature engineering we can use hypothesis testing (p values) 👍
Sir, you are a life saver, this is very helpful ❤️
Yes waiting for this one
I just went into deep learning into this particular algorithm. It seems interesting
Thanku so much, helpful n interesting video
Really Excellent content. Please do provide for all machine learning algorithm. It will be really helpfull
Awesome video krish.. Pls upload more video related to data science and machine learning
Great series
Krish sir...if someone has accounting experience .....how much hours/months required to learn AI which is required for accounting & finance field???
Krish is the shit man! Awesome stuff! 👍
thank you once again sir
Eagerly waiting for this video.
@yashvaibhav925
3 жыл бұрын
So we have two Yash eagerly waiting
Linear regression does not make any assumptions about the distributions of the dependent and independent variable to be normally distributed. I believe Point 4 in assumption is incorrect because most of the online resources does not mention this.
@fahidlatheef564
3 жыл бұрын
I also thought so. However, the normality of residuals is mentioned in a few sources.
@cefax875
3 жыл бұрын
Yes correct.But parameter estimation is based on the minimizing the squared error, few extreme observations can have a disproportionate influence on parameter estimates.So Normality is essential.
@DeepROde
3 жыл бұрын
You're right. Simplest example would be straight line, which is not normally distributed. Actually the errors (residuals) should be iid & normally distributed with 0 mean & finite variance. If this assumption is satisfied, square error loss function will give maximum liklihood estimation of weight vector.
@DeepROde
3 жыл бұрын
However, I must add that saying Y|X is normally distributed will be correct because it follows from the assumption that residuals are normally distributed (maybe that's what he meant when he wrote 'fixed value of X'.
Where is those links of videos ,that you told in this video
hello krish sir, for linearity why mean of Y is considered against X why not directly X and Y?
how to access the link given in the video???
where do i get those youtube links
The assumption of linear regression mainly talks about the error terms and is not related to X and Y. Could you please confirm when you refer to the transformation(log transformation) is required to make it normally distributed. was it for error terms or X & Y?
How to copy links which are provided in Vedio ?
where I can get that Arja Basu's video?
Can someone please answer how to approach this situation in regression : The target variable is distributed in a biased manner(50% in 0-300 and 30% in 300-500 and 10% in remaining) , how will you approach such scenario?
Thank you 👍
Thanks Krish
yeh kyu band kar diya sir???? please continue this preparation..........
Thanks you sir
If you're using the analytical solution(OLS regression , instead of gradient descent algorithm) for creating the Linear Regression Model , feature scaling won't be of much use.
@aryankaushik3761
2 жыл бұрын
Ols is old method now . You can learn but it's not used extensively.
9:41 was a very crucial point.
Where i find these links
Please cover all the loss functions in a separate video
Why observations need to be independent of each other? Any specific reason for this assumption?
@sandeelg_lite
3 жыл бұрын
if next observation depends on current observation it comes into time series analysis. For that another kind of regression is used.
sir, in the video you have mentioned that Feature Scaling is required for linear regression but i didn't find any improvement in performance before and after scaling. please suggest ..
@krishnaik06
3 жыл бұрын
It improves the performance while training
@gnaveen451
3 жыл бұрын
@@krishnaik06 Thank you sir, I will check in that way..
Good videos Krish. Perhaps you must minimize the repetitive use of words. This will have several benefits. 1. minimize your efforts, 2. not being boring for the listener, 3. verbal information content will be more concise, 4. the videos will be short in length and save time for the listener.
Wow wow wow
Should i practice by writing my own full code including the hypothesis functions, cost functions, gradient descent or fully use sklearn?
@kesavae9552
3 жыл бұрын
If you know how write them and you are confident, then go ahead with sklearn and save time. If you want to know how to write them write everything from scratch.
@KnowledgeAmplifier1
3 жыл бұрын
You can try one time to get complete feeling and complete understanding in working of the algorithm and next onwards that lifetime experience will help you in working with built-in functions :-)
Lots of improvement is needed. Construction and gradient descent is not converted
thanks sir...one special request from my side plz sir added more n more questions which are mostly asked by the interviewers....thanks sir
@krishnaik06
3 жыл бұрын
These are all the possible questions
@anshusingh536
3 жыл бұрын
@@krishnaik06 ok....thanks a lot sir.....its help me a lot
It is not that we haven't studied but we don't have been taught
I wish you could just jump straight into and focus on the main thing! Thanks though.
I wonder how guys like Krish, striver, and Kunal are providing a 100-dollar course completely for free.. and the key is it's enough to pay for youtube than for Spotify, or some other course-providing platform, or fitness app. Every thing is available for completely free in this internet era. we must find the correct resource!
Krish, How to be member and access member-reserved material?
@rohitchan007
3 жыл бұрын
Join the membership
@rwagataraka
3 жыл бұрын
@@rohitchan007 How? Just subscription?
@rohitchan007
3 жыл бұрын
Go to his KZread page and there you will find an "Join" click on that and select the appropriate plan that suits you and pay the amount. You will become the member.
Isn't the Normality assumption only for the residuals?
@weiyang2116
3 жыл бұрын
I believe the case..and there's a common misconception that normality applicable for dep. var, which is false
Please make sure you are talking to the point!
#1video1algo
「コンテンツを調整する必要があります」、
You can do the videos being less dramatic