AdaBoost, PCA (Capstone Project)

Capstone Project: PCA & AdaBoost concepts are applied to 'Car Detection' from images

Dataset: UIUC Image Database for Car Detection ( https://cogcomp.cs.illinois.edu/Data/Car/ )
PCA
(a) Finding the best k, where k is the dimension of the optimal subspace to which the data is projected.
(b) Suitable classification algorithm on new data and various performance measure.

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
| AdaBoost : Implemented in 2-dimensional projection space. (i.e.Number of Pricipal Components = 2) |
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
AdaBoost :
AdaBoost (Adaptive Boosting) generates a sequence of hypothesis and combines them with weights.

::Choosen Weak classifiers::
1. GDA
2. Knn (NumNeighbors = 30)
3. Naive Bayes
4. Linear (Logistic Regression*)
Refer to: https://www.iist.ac.in/sites/default/files/people/in12167/adaboost.pdf

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Contents

Initialization, Dataset : *'CarPixels.csv'* :: Generated from: UIUC Image Database for Car Detection
Sample Images (Random)
Applying PCA
Performance Measure & Optimal number of Principal Components (K)
Explaind-Variance Curve
Performace Measure
Reconstruction of Images

- - - - - - - - - - - - - - - - - - - - - - - - - - - - -
| Adaboost (GDA, Knn, NB, Logistic) |
- - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Applying AdaBoost
Initialization (2-dimension)
Gaussian Discriminant Analysis Classification
Knn Classification
Naive Bayes Classification
Logistic Regression
Conclusions

Related Examples:
1. SVM
https://in.mathworks.com/matlabcentral/fileexchange/63158-support-vector-machine

2. SVM using various kernels
https://in.mathworks.com/matlabcentral/fileexchange/63033-svm-using-various-kernels

3. SVM for nonlinear classification
https://in.mathworks.com/matlabcentral/fileexchange/63024-svm-for-nonlinear-classification

4. SMO
https://in.mathworks.com/matlabcentral/fileexchange/63100-smo--sequential-minimal-optimization-