🤔QUESTION OF THE DAY: How do you handle missing data? Comments down below! 😃 💗Help support this RU-vid channel by hitting the Subscribe button, Like button and Comment down below! 👇
Marco Festugato It’s a pleasure, more videos on Data pre-processing is in the making. Please stay tune for more. P.S. I also referenced your name in the video description
@@DataProfessor thank you very much, I appreciated it a lot! I've started studying R, there should be more youtube channels like yours, I really like it. Don't stop please! And consider writing me if you come to visit Italy :)
Sir, can you make a another video on missing data of numerical type where some advanced techniques applicable when mean and median is not working. You are write thing, keep it up and make more video on R.
Hi Deep patel, you got it, next episodes of the ‘Data pre-processing series’ will have more advanced topic of handling missing data. Please subscribe and hit notification bell for update of latest video.
Can you do a video regarding some of the important algorithms to master, like regression, classification etc? Trying to find information online can sometimes be overwhelming because they throw a million things at you at once. Thanks in advance.
Hi Bazi, thanks for your suggestion. I'll put this into the to-do list. In the meantime, please subscribe and hit the notification bell for updates of new videos.
Hi Bazi, following up on your suggestion, I've just released a video Data Science 101: Overview of Machine Learning Model Building Process at ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-BOk1hlCPW0c.html I also referenced your name in the video description, check it out!
Once again..Amazing video!!. However, at the end of the video you suggested to upload other videos on how to deal with missing values for other types of data which are not numerical such as factor, categorical , ordinal..Are you still going to upload these videos? Thanks
Thanks Desmond for the comment and kind words. Yes, definitely and thanks for reminding, I will put this in the priority to-do list. If there is any other topic you would like to see, feel free to suggest.
Great question! There's a mice R package that allows imputing both missing quantitative and qualitative values, the original paper describing the package is available at www.jstatsoft.org/article/view/v045i03
Hello Professor. I have a question which is related to missing data. I have a dataset cars_missing and there has two missing data. When I run sum(is.na(cars_missing)), it's only show 1. Also, When I run view(cars_missing), I can see in column "cubicinches" there has "NA", but in column "brand" there only show blank without NA. I think that's why when I run sum(is.na) there only show 1. Can you explain why?
really nice video . but sir i am getting error while loading data Error in function (type, msg, asError = TRUE) : error:1407742E:SSL routines:SSL23_GET_SERVER_HELLO:tlsv1 alert protocol version
It is difficult to say which is better than the other. If using mean for imputation, the mean value is preserved, meaning that the new resulting mean value would not be changed while this would cause perturbation to the median value; and vice versa for using median imputation. It should also be mentioned that either mean or median imputation ignores the relationship with other variables of the dataset. A better approach would be to perform regression imputation which would consider the relationship with other variables of the dataset. Further complicating this issue is that this discussion is valid for missing values that are numeric. If they are categorical then mean/media imputation would not be valid, thus favoring the logistic regression for predicting the categorical missing values (in a similar fashion to regression imputation for predicting numerical missing values).
Hi, for imputing categorical data, I would recommend to use the mice R package. Please kindly refer to this for more information www.coursera.org/lecture/missing-data/mice-r-package-N6STE
Brilliant video! Clear and simple explanations for those just starting out in this area :) Did you ever make videos on how to deal with ordinal/categorical missing data? I can't seem to find them on your channel?