[https://www.codecademy.com/journeys/data-scientist-ml/paths/dsmlcj-22-machine-learning-foundations/tracks/dsmlcj-22-machine-learning-fun/modules/eda-machine-learning-models-53f6f860-50fe-40e0-a6aa-1268771d0bcb/articles/eda-prior-to-unsupervised-clustering](https://EDA for Machine Learning Models)
In this article there’s a section that says “Feature reduction for EDA”
The article states shows a data frame and then states:
“We are looking for the highest weighted feature for each component of interest. For component 1, this would be
Flavanoids , component 2,
Color_Intensity , and so on. These may be particularly important features to include in our model.”
Im not sure how component 2, ‘Color_Intensity’ has the highest weighted feature the number for this is -0.529996 to me a negative number would be the lowest weighted, especially since there are several more positive weights in the component 2 column.
I also thought, well maybe they mean looking at the Color_Intensity row but there are more positive numbers in the row as well. I guess it is lower than the positive number is high. Is that what this is saying?
I guess Im not sure what the weight of a feature is as it has not been previously discussed.