Looking at Principal Component Analysis Loading Scores in Python
In this follow-up video to my big Principal Component Analysis (PCA) video, I show you how to analyze the results of PCA by showing you how to look at the loading scores in Python and explain how to read the results to discover what features/variables hold the highest amount of weight for the Principal Component and what that means for feature selection/reduction of models. I go over this for both: the Iris Flower PCA example, and the Crypto PCA example from my previous video.
Unrelated, I also go over how to read a confusion matrix for a multi-class classification model (in this example, the 3-class Iris Flower Classification Model from a previous video). I've been meaning to go over this for a while, and decided to throw it in here too. Finally I also review how to shrink/change the size of markers for plotly graphs, something that bothered me in the last video!
Useful Reference(s):
Scikit-Learn Documentation on the Principal Component Analysis: https://scikit-learn.org/stable/modul...
Scikit-Learn PCA Example on the Iris Dataset: https://scikit-learn.org/stable/auto_...
Plotly Graphing Library: PCA Visualization in Python: https://plotly.com/python/pca-visuali...
The Crypto Dataset: https://www.kaggle.com/odins0n/top-50...
Previous Video on PCA: • Principal Component Analysis in Python
Credits where Credits are Due:
-Subscribe End Screen: Graphicki • Video
-Live Animation: Motion Chroma | Leader • Livestream Intro [Free to Use] ♥
-Music: Song: Without You, Artist: MNRVA, / ximertracks
-Streaming Software: OBS
Thank you for watching!
Timestamps:
00:00 PCA
19:23 Confusion Matrix for Multi-class classification