Shape Classification and the FFT.


Problem 1
Build an appropriate training set for the classification of the four shapes in the time domain and in the frequency domain (using the Fourier Transform). Your training set will include clean and noisy shapes as described below using a random phase to build each case. The table indicates the number of training cases for each shape.

Shape    Clean    10% Noise    20% Noise    30% Noise  
Cardioid250250250250
Rose150150150150
Irish200200200200
Heart150150150150

Classification

Hint
To keep your files organized use the file names as described in the table below.

Test    Time Domain    Frequency Domain  
Build Train SetBuidTrainSetT.lab BuildTrainSetF.lab
Training SetTrainSetT.csvTrainSetF.csv
Build Validation SetBuidValidSetT.csvBuildValidSetF.csv
Validation SetValidSetT.csvValidSetF.csv
TrainingTrainT.labTrainF.lab
Check TrainingCheckTrainingT.labCheckTrainingF.lab
(Trainining) Confusion MatrixtrainConfT.emftrainConfF.emf
Validation ValidationT.lab ValidationF.lab
(Validation) Confusion MatrixvalidConfT.emfvalidConfF.emf

Problem 2
Build an appropriate validation set for problem 1 using random phase. The table below describes the number of validation cases for each shape.

Shape    Clean    10% Noise    20% Noise    30% Noise  
Cardioid100100100100
Rose50505050
Irish40404040
Heart90909090

Problem 3
Design and train an ANN for the classification of the shapes.

Problem 4
Check the training: (a) Compute the confusion matrix using the training set. (b) Save the confusion matrix as vector image.

Problem 5
Perform the validation of the ANN. (a) Compute the confusion matrix using the validation set. (b) Save the confusion matrix as vector image.

Problem 6
Generate a report in Microsoft Word. Write some conclusions in the report trying to compare your results using data in the time domain with the data in the frequency domain.

© Copyright 2000-2019 Wintempla selo. All Rights Reserved. Sep 05 2019. Home