There are various types of normalization techniques as mentioned in the link below :- en.m.wikipedia.org/wiki/Norma... In this video we will only be focusing on the Z-Score
If I have a data set which is coming every 5 minutes, in this case the mean and standard deviation is varying dynamically, how to calculate Z Score for this dataset
Good video sir. But its often confusing that when to use Z score and when to use normalization for scaling between 0&1. What if data is not following bell shape? Please kindly explain😊
Hi Thank you for your comment. Whilst I haven't explained the calculation of Mean and SD in the video. I have created two other separate videos for the same. Please find below the link for the calculation of the Mean and other central tendencies :- th-cam.com/video/qclooEL0LPg/w-d-xo.html Furthermore please find below another link for the calculation of Standard Deviation (SD) :- th-cam.com/video/b_qbCjGvO_Q/w-d-xo.html Hope this helps! I have created a ML playlist (th-cam.com/play/PL_jahYUFcXYyev6tw-rhjpYQ6FvFVtD42.html) where all my videos are in sequential order whenever convenient please have a look. In that playlist the videos for Mean and SD come before this video it would be easier for you to navigate. Thanks!
Thank you for your comment this was because in Excel (when I did a quick calculation) I forgot to add decimal points it was rounded. In any case you are right .. But overall it wouldn't affect the scale hugely. I will make sure this doesn't happen again
Hi, thank you for watching and for pointing out your concerns. However, your claim is not entirely accurate as Normalization can be achieved using 2 different formulas as can be identified in the following link: www.codecademy.com/articles/normalization#:~:text=Z-Score%20Normalization,-Z-score%20normalization&text=If%20a%20value%20is%20exactly,will%20be%20a%20positive%20number.&text=This%20is%20the%20same%20data,re%20using%20z-score%20normalization. Deploying Z-Score as I have done in the video is also referred to as normalisation. The alternate formula you seem to be referring to is called Min-Max Normalisation and is also a valid system, however, it is not the only available method for normalisation. Anyhow, in order to avoid any further confusion for my viewers, I will do the necessary changes and rename it as Z-Score, since though Normalisation and Standardisation are often used interchangeably and this can lead to some uncertainty, in fact standardization is a subset of normalization as you can see in the following link :- en.m.wikipedia.org/wiki/Normalization_(statistics). The specific terminology of Min-Max and Z-Score are more distinct. My intention has never been to teach ‘wrong things’, however, I understand the cause of confusion in this instance. I apologize for any inconvenience caused that led you to misconstrue my video. Furthermore, whilst your comments are extremely valuable to my channel’s growth, your insinuation is not appreciated, and I would welcome a tone of constructive criticism rather than barbed insinuations in future comments.
The clearest explanation till now! Keep it up.
Overwhelming explanation. Looking forward to seeing more videos based on the ML-R integration approach.
thanks a lot, I've been looking for that the entire day!
Thx for the video!! Very well explained :)
If I have a data set which is coming every 5 minutes, in this case the mean and standard deviation is varying dynamically, how to calculate Z Score for this dataset
Good video sir.
But its often confusing that when to use Z score and when to use normalization for scaling between 0&1. What if data is not following bell shape? Please kindly explain😊
Thanks!
Thank you for the knowledge shared, please explian/share video how to find our Mean and SD for the same values shown in this video. Thank you.
Hi Thank you for your comment. Whilst I haven't explained the calculation of Mean and SD in the video. I have created two other separate videos for the same.
Please find below the link for the calculation of the Mean and other central tendencies :- th-cam.com/video/qclooEL0LPg/w-d-xo.html
Furthermore please find below another link for the calculation of Standard Deviation (SD) :-
th-cam.com/video/b_qbCjGvO_Q/w-d-xo.html
Hope this helps!
I have created a ML playlist (th-cam.com/play/PL_jahYUFcXYyev6tw-rhjpYQ6FvFVtD42.html) where all my videos are in sequential order whenever convenient please have a look. In that playlist the videos for Mean and SD come before this video it would be easier for you to navigate. Thanks!
Thank you so much 🙏🏻
What is the difference between Normalization and standardization
Equation of Normalization:-
X' = (X - X_min) / (X_max - X_min)
Here scale = [0,1]
Equation of Standardization:-
X' = (X - X_mean)/(X_standard_deviation)
i found here that the mean of the height = 167.4 .. and sd = 2.4 why we put 167 and 3 instead
Thank you for your comment this was because in Excel (when I did a quick calculation) I forgot to add decimal points it was rounded. In any case you are right .. But overall it wouldn't affect the scale hugely. I will make sure this doesn't happen again
It's standardization, not normalization. Normalization has a different formula. Stop teaching wrong things.
Hi, thank you for watching and for pointing out your concerns. However, your claim is not entirely accurate as Normalization can be achieved using 2 different formulas as can be identified in the following link:
www.codecademy.com/articles/normalization#:~:text=Z-Score%20Normalization,-Z-score%20normalization&text=If%20a%20value%20is%20exactly,will%20be%20a%20positive%20number.&text=This%20is%20the%20same%20data,re%20using%20z-score%20normalization.
Deploying Z-Score as I have done in the video is also referred to as normalisation. The alternate formula you seem to be referring to is called Min-Max Normalisation and is also a valid system, however, it is not the only available method for normalisation.
Anyhow, in order to avoid any further confusion for my viewers, I will do the necessary changes and rename it as Z-Score, since though Normalisation and Standardisation are often used interchangeably and this can lead to some uncertainty, in fact standardization is a subset of normalization as you can see in the following link :- en.m.wikipedia.org/wiki/Normalization_(statistics). The specific terminology of Min-Max and Z-Score are more distinct.
My intention has never been to teach ‘wrong things’, however, I understand the cause of confusion in this instance. I apologize for any inconvenience caused that led you to misconstrue my video. Furthermore, whilst your comments are extremely valuable to my channel’s growth, your insinuation is not appreciated, and I would welcome a tone of constructive criticism rather than barbed insinuations in future comments.