The sequence of steps for designing a machine learning algorithm are:
1. Define a Gradient Descent Algorithm
2. Iterate the gradient descent algorithm using computer simulation methods.
3. Compute the derivative of the discrepancy function with respect to the parameters of the learning machine.
4. Design the discrepancy function to evaluate the predictions of the learning machine.
5. Decide if the environment provides the learning machine with explicit feedback regarding what is the right answer for a given prediction.
6. Define feature vectors for representing events in the environment. Decide if actions of the learning machine modify the environment.
7. Evaluate the performance of the algorithm on both training data and test data.
sequencelength = length(mysequence)
disp('STEPS FOR MACHINE LEARNING ALGORITHM DESIGN')
for i = 1:sequencelength
disp(['Step#', num2str(i), ':', proceduresteps{mysequence(i)}])
end