Data Mining - Principal Component Analysis (PCA) and Multidimensional Scaling (MDS) in 7 MINUTES

preview_player
Показать описание
This video gives the key takeaways on the Curse of Dimensionality and two dimension reduction techniques: Principal Component Analysis and Multidimensional Scaling. Dimension reduction techniques can be used to tell us which variables in our data are most important. Principal Component Analysis (PCA) and Multidimensional Scaling (MDS) are useful data analysis techniques that every Statistican and Data Scientist ought to know.

In this video, we go over the main ideas behind PCA and MDS in order to give an intuitive sense of the methods.

0:00 - The Curse of Dimensionality (COD) and "Dimension Reduction"
1:05 - Principal Components Analysis (PCA)
3:11 - Multidimensional Scaling (MDS)
Рекомендации по теме
Комментарии
Автор

Very clear and simple in its terms. I am truly grateful.

ammarabdelaal
Автор

Very good ! Congratulations !

I've seen many conceptual presentations of PCA, but yours was the best one.

HiltonFernandes
Автор

wow thank you for this clear explanation! the rotation of axis vs eigenvector part is concise and effective.

weinansun
Автор

The clarity of your presentations is both admirable and easy to follow. Thank you for sharing!

willykitheka
Автор

Please do some examples on real data.
By the way appreciated your efforts

thallkhan
Автор

I don't understand how and why the first component explains the most important variation in the data. And do you calculate the percentages of each one?

skyalmillegra