Contents
Does LDA assume normality?
LDA: multivariate normal with equal covariance That is, within each class the features have multivariate normal distribution with center depending on the class and common covariance Σ. The probabilities P(Y=k) are estimated by the fraction of training samples of class k.
Can LDA be used for dimensionality reduction?
Linear Discriminant Analysis, or LDA for short, is a predictive modeling algorithm for multi-class classification. It can also be used as a dimensionality reduction technique, providing a projection of a training dataset that best separates the examples by their assigned class.
What are the criteria used by LDA for performing classification?
Two criteria are used by LDA to create a new axis: Maximize the distance between means of the two classes. Minimize the variation within each class.
Is the LDA algorithm a dimensionality reduction?
In this post, I am going to continue discussing this subject, but now, talking about Linear Discriminant Analysis ( LDA ) algorithm. LDA is defined as a dimensionality reduction technique by authors, however some sources explain that LDA actually works as a linear classifier.
Why do we use linear discriminant analysis in LDA?
With or without data normality assumption, we can arrive at the same LDA features, which explains its robustness. Linear discriminant analysis is used as a tool for classification, dimension reduction, and data visualization.
How is linear discriminant analysis used in dimension reduction?
Linear discriminant analysis is not just a dimension reduction tool, but also a robust classification method. With or without data normality assumption, we can arrive at the same LDA features, which explains its robustness. Linear discriminant analysis is used as a tool for classification, dimension reduction, and data visualization.
When to use LDA for non linear separability?
However, if the dataset is not linear separable the LDA will try to organize your dataset in another space as the maximum linearly separability as possible, but it still be examples overlapping between classes because of non-linearly characteristic of data.