Hi sir! Thank you for your video, it's really helpful. I've one question, may I know what python version you use for installing misspingpy packages? I've tried to install missipinpy packages but it doesn't work (got an error).
Thanks for sharing. Missing values is really a critical problem in data engineering which needs to be addressed as early as possible in the analysis chain.
Awesome Aman. As usual, you have an art of delivering content in an effective way. Brilliant! I often get confused which imputation works best. In your example, how to infer, which technique suits the problem statement? Can you please provide some clarification?
Thanks Harish, one way of looking is more you understand your missing data better call u can take for example, multiple imputation (MICE) works well with MAR and MNAR types. If I don't understand domain much, I ll go with probably mean/median/mode way. If I know my data is kind of high variance, I can go for missforest. All these are few things we can check. If u ask me in real World use cases, We try to see various methods which suits well with
Hello Aman ! Tanks as usually for your great videos. I have a question, what about replacing listing values by out of range ones as -999 ? I'm dealing with a dataset where values are missing bc these are hard to record and I thought to replace those by -999 (that's a dataset with soccer games results)
Hi Aman, with no intentions to trouble you, can you pls mention the folder name in the google drive where I can find the .pynb file for missing value imputation ..thanks
Aman You are awesome. I am post graduate student at IIT Delhi and believe me you explain in much better way than my professor.
Thanks Piush, your comments motivate me.
Hi sir! Thank you for your video, it's really helpful. I've one question, may I know what python version you use for installing misspingpy packages? I've tried to install missipinpy packages but it doesn't work (got an error).
Thanks for sharing. Missing values is really a critical problem in data engineering which needs to be addressed as early as possible in the analysis chain.
I can bet that this is one of the best videos on missing value handling available in TH-cam. Great work appreciated 👍
Thanks Sourav. Love your comments always :)
Thank you so much for the great lesson. Greetings from Korea!
Hi Aman. Thanks for the upload. You just gained a subscriber.
Thanks Yusuf
Nice
Koti koti pranam 🙏. Was eagerly waiting for this.
Thanks, do tje assignment in the end 😇
@@UnfoldDataScience sure
Very nicely explained. Also if possible, it would be great if you could share the working notebooks as well.
awsm video
Thanks Mukesh. Happy holi
Awesome Aman. As usual, you have an art of delivering content in an effective way. Brilliant!
I often get confused which imputation works best. In your example, how to infer, which technique suits the problem statement? Can you please provide some clarification?
Thanks Harish, one way of looking is more you understand your missing data better call u can take for example, multiple imputation (MICE) works well with MAR and MNAR types.
If I don't understand domain much, I ll go with probably mean/median/mode way.
If I know my data is kind of high variance, I can go for missforest.
All these are few things we can check.
If u ask me in real World use cases, We try to see various methods which suits well with
@@UnfoldDataScience Thanks for providing clarity. I'm in better shape now. Appreciate your quick response.
Thanks
good video sir
Thanks Mohit
Pls can you make a videos on A/B Testing. Quiet an important topics
Sir , if you could teach these imputation techniques Probabilistic Matrix factorization and Bayesian maximum entropy as they are the state of the art.
Any tips of how to integrate categorical data? Using the labelencoder to encode the features creates some issues
miceforest and missforest can handle categorical data?
Hello Aman !
Tanks as usually for your great videos.
I have a question, what about replacing listing values by out of range ones as -999 ? I'm dealing with a dataset where values are missing bc these are hard to record and I thought to replace those by -999 (that's a dataset with soccer games results)
thank you for making this video sir!
Welcome Shubham, notebook in my google drive. Link in description take it and try deck imputation, paste your code once u do 🙂
Thanks
plz make a video how to remove feature using backward approach to reduce feature in python
Please search for "recursive feature elomination unfold data science" On TH-cam
Thank you
Glad you found it helpful
strategy is hyperparamater 👨💻which guide the paramater "imputer".
🙂
what about if categorical values are missing
Hi Aman, with no intentions to trouble you, can you pls mention the folder name in the google drive where I can find the .pynb file for missing value imputation ..thanks
No folder, it's outside
share repo of notebook
It's in my google drive as always Subhash. Link in description
@@UnfoldDataScience what is name of notebook?
@@subhashachutha7413 Python code - imputation techniques
@@UnfoldDataScience thank you
@@UnfoldDataScience can't find the link in the description sir