Project work on LDA classification
Why to use LDA? We LDA to reduce detention(features). In the real world there will be many features. To reduce the number of features and preserve the discriminatory information.
In LDA we need to follow step by step process to find the object class Step 1: Extract the data from Iris Step 2: Calculate the mean for each class and each feature seperately Step 3: find the inClass scatter matrix Step 4: find the interclass scatter matrix step 5:
InClass scatter matrix we need to use Sw = Σ(i=1 to c) Σ(x in class i) (x - μi)(x - μi)^T for finding Inclass Scatter matrix.
Sw meant scatter matrix c is the number of classes x is the sample in class i μi is the mean of the class
BetweenClass scatter matrix
we need to use SB = Σ(Ni * (μi - μ) * (μi - μ)^T) for finding Between Class Scatter matrix
Ni is the number of samples in a class μi mean of class i