Basic Supervised Learning Algorithm
Discriminant function analysis is a statistical analysis to predict a categorical dependent variable (called a grouping variable) by one or more continuous or binary independent variables (called predictor variables). Discriminant analysis works by creating one or more linear combinations of predictors, creating a new latent variable for each function. These functions are called discriminant functions. - Wikipedia.
 |
| Supervised Learning |
Steps to predict the class:
- For given test samples extract vector f.
- Evaluate Di(f) for 1 ≤ i ≤ k.
- The class which is predicted is the class with highest D(f) value.
Let us see small pseudo-code for taking decision and applying learning rule
Initialize the coefficients randomly.
while( ! converged ){
for every training sample{
predict class ( say j )
let actual class j'
if(j==j')
Do Nothing
else
apply learning rule and increment error
}
check for convergence
}
while( ! converged ){
for every training sample{
predict class ( say j )
let actual class j'
if(j==j')
Do Nothing
else
apply learning rule and increment error
}
check for convergence
}
Let us see simple MATLAB code to classify the flowers in IRIS dataset
load fisheriris;
gscatter(meas(:,1), meas(:,2), species,'rgb','osd');
xlabel('Sepal length');
ylabel('Sepal width');
We'll explore the code in other tutorial (Basic Classification using MATLAB). gscatter(meas(:,1), meas(:,2), species,'rgb','osd');
xlabel('Sepal length');
ylabel('Sepal width');
The output of this MATLAB code shows how the data from dataset is classified as different flowers.
Subscribe to:
Post Comments
(
Atom
)
No comments :
Post a Comment