Live-Feature Engineering-All Techniques To Handle Missing Values- Day 3
Dataset link: drive.google.com/file/d/1hJgt...
Live Streaming Playlist: • Live stream playlist
Telegram link: t.me/joinchat/N77M7xRvYUd403D...
github link: github.com/krishnaik06/Featur...
Join the Ineuron Affordable course
ineuron1.viewpage.co/Deep-lea...
Please donate if you want to support the channel
Gpay: krishnaik06@okicici
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 too
/ @krishnaikhindi
Connect with me here:
Twitter: / krishnaik06
Facebook: / krishnaik06
instagram: / krishnaik06
Пікірлер: 55
Thanks sir for this series, no one is teaching such techniques , it's really helpful for everyone 🙏
Commendable job, people can record and do coding offline but to do it on live streaming is great
Thank You so much! I really enjoyed this session.
#respect.... thanks for sharing your knowledge...!!!
This is perfect sir plz go ahead by this kind of feature engineering I truly learn a lot ...plz continue sir 💖💖😊 session is going awesome
god bless you krish, you rock
Thanks a lot sir, awesome explanation and hats off for your patience and dedication even after your regular office work
Sir you are taking so much pain to teach the best of your knowledge . Thanks a lot sir.
Thank you alot best teacher ever
thank you so much sir explanation by you is very easy to understand...... :)
From 200K while recording this video to today4/7/2022 - 600K subscribers, kudos to you Krish for your effort
Very useful sir!! Thank you
Great class!
perfect explanation.
very useful sir Thanks...
Thank you so much sir
Nice explanation 💟
thank you sir
We can use embeddings for features with large category which encodes the feature in small feature space
Very useful sir😍😍😍
Finished watching
Yes
Sir can we use technique like filling the missing values either by back value or next value ......
Super useful 🤟🏻
@amponsahwellington3197
3 жыл бұрын
Naik, I sent you a mail . Pleae reply me
hey guys, does feature engg require knowledge of ML as well? or can u plz tell the pre requisites for FE?
god bless this man lol
How do we understand which are my categorical variables from a dataset?
#KingKrish
krish sir please record the video and upload it !!
why are we converting the data set of mercedesbenz into list??
but we have to do the same for other categories too still we may get 5x10 = 50 more columns right which is expected
@nitikeshsaini7655
11 ай бұрын
Right
Can anyone plz tell about the code .. For category in lst_10: Df[category]=no.where.......
Can you explain the winners solution of machinehack is that possible
I have bought your Membership Krish Sir. I Saw that it unlocked the Projects Playlist, but if you have any other material than How do I access the Data Science Material?
@krishnaik06
3 жыл бұрын
Hi Harsh please check the community post. All the info is given there
@harshmakwana8001
3 жыл бұрын
@@krishnaik06 Thank you sir!☺️
one question is when we have 100 of categorical values in the categorical variable and when we apply the method of selecting the top 10 of them and then we perform the encoding, whether we will drop that original categorical variable from our dataset?
@joeljoseph26
5 ай бұрын
check KDD Orange CUP competition from kaggle. He has also covered a video about picking the top 10 features and ignoring the next. You can also try, mean-encoding,freq/count encoding, target-guided(if you need rank for the lables)
Part 2- Handling Categorical features starts at 1:12:0
@rebeccakipanga478
11 ай бұрын
Thank you
Hey Krish, Do we have any one method for replacing NAN values, which will work for all types of datasets?
@zshan101992
3 жыл бұрын
I guess there is no such all in one method.. usually impution with mean median mode is preferred to get complete data fast butit impacts correlation also and distort the variance so choose accordingly.
can anyone give the code to convert all other features apart from x1
Most frequent comming category means (mode) , why we can't directly imputing na by mode ???????? like this df.bsmatq.fillna(df.basmqt.mode())
Sir! is it ok to replace null values(0) with imputation?
@bruhm0ment767
Жыл бұрын
If the number of rows with null values is less compared to the total dataset, the rows can simply be dropped. However, if the number of rows with nulls is high we need to impute
wanna join you but technical hurdle while payment,
finished practice coding
If 47% of the data is missing in a column, does it make sense to actually impute the data? Would it not be better to delete the column itself, if almost half the values are synthetic?
Naik, I sent you an email. Pleae check and reply when you are not busy.
i am literally killed... please Krish, stop smacking your mouth...
thank you sir