How did you find this video? Please let me know in the comments below!
@ngocuyennguyen4687
Жыл бұрын
I came upon this by searching for ways to improve clustering result.
@Alisq26
3 жыл бұрын
Thank you so much, you`re a legend. All parts were very informative!
@YiannisPi
3 жыл бұрын
Glad it helped
@Asylum_M
4 жыл бұрын
Hello Yiannis, thank you for this video series. I can say I've enjoyed your explanation, but I have some questions. As far as I've learned, K-Means algorithm is rather strict towards the input data: it must be normalized. Why did you skip this step in your EDA?
@brandonruiz1746
2 жыл бұрын
awesome breakdown brother thanks for this. I appreciate the detailed breakdown.
@LuisCamacho-ql5so
2 жыл бұрын
Great videos
@MrKapilsingh
3 жыл бұрын
You are amazing! I understood each and everything you explained. Great job!!!
@LaurentD90
3 жыл бұрын
Very good. Thank you
@dudee420
4 жыл бұрын
Hi Yiannis I really like your videos as they are very learning and you explain the things very easily. Also as you are working as data scientist in some firm, can you give an example overview practically what exactly is done in normal routine task in this profile off course i dont want you to showcase the company data.. just want a real based example as a data scientist profile
@YiannisPi
4 жыл бұрын
Hey Manish, glad you like the videos! You mean more of a "vlog" ? I do have a few videos explaining what does a data scientist do day to day if you check my channel
@YiannisPi
4 жыл бұрын
@wise guy This is a real life scenario. Every single company wants to segment their customers for many reasons. Hence this is applicable to any business. If you watch the video I will be uploading today which is deployment and insights, you will understand better!
@amospeter7772
9 ай бұрын
Hello Brother. Can I know how do you know the clusters location since you run PCA. From my understanding, PCA will create a new set of variables and once you run clustering again from PCA, how do you determine the clusters is located at a specific row since the PCA is new variables?
@yazanaloqaily5476
4 жыл бұрын
Thanks for this perfect video Please could you provide me any resource that solves examples about Gaussian Maximum likelihood?
@atb0007
2 жыл бұрын
HI YP when will you post the v4- dashboard part?
@go1chase1the1sun1set
3 жыл бұрын
I have a full categorical data set, at first I used K-Modes but could not get a scatter plot visualisation, so I one hot encoded all of them and did KMeans and Im trying to get a scatter plot, is it possible?
@SATULAL
Жыл бұрын
How did you end up solving this issue? I currently have a dataset with categorical variables that I converted to binary 0/1 variables though I am still unsure if that was necessary to do so. I have household demographics of age dummies, employment status, education, income below median dummy, state dummies and city size dummies that I want to cluster. Through my online searches, the recommendations I got were to use: a) K-means but with a different distance metric (Jaccard distance) but this depends on the dimensionality of the data. If it is super large, this could be an issue b) K-modes or K-prototype as you have used above I'd love to hear what you ended up doing!
@nazaninmashayekh3148
4 жыл бұрын
Thank you for your perfect video. I think your data set does not have ground truth. You did tell nothing about clustering evaluation in your videos. Can we use the silhouette score function or something else to evaluate the clustering result while our data set does not have a ground truth column?
@YiannisPi
4 жыл бұрын
Check the V2 video for evaluation. I talk about it there :)
@simonihegbu8388
2 жыл бұрын
Hey man, I ran into problems with my data when doing K-means clustering, how can I show you my model maybe you can check it?
Пікірлер: 20