Scribd Logo
Upload

Welcome to Scribd!

  • 9 views

    Principle Component Analysi12

    Uploaded by

    neelamysr
    Principle Component Analysis (PCA) is an algorithm for dimensionality reduction and feature extraction. It transforms a dataset onto a new feature space of lower dimension while retaining as much information as possible. The steps of PCA involve getting data, mean normalization, calculating the covariance matrix, finding eigenvectors and eigenvalues of the covariance matrix, choosing principal components to form a new feature vector, and deriving the transformed dataset.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Principle Component Analysi12

    Uploaded by

    neelamysr
    9 views5 pages
    Principle Component Analysis (PCA) is an algorithm for dimensionality reduction and feature extraction. It transforms a dataset onto a new feature space of lower dimension while retaining as much information as possible. The steps of PCA involve getting data, mean normalization, calculating the covariance matrix, finding eigenvectors and eigenvalues of the covariance matrix, choosing principal components to form a new feature vector, and deriving the transformed dataset.

    Original Description:

    PCA for Ml

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Principle Component Analysis (PCA) is an algorithm for dimensionality reduction and feature extraction. It transforms a dataset onto a new feature space of lower dimension while retaining as much information as possible. The steps of PCA involve getting data, mean normalization, calculating the covariance matrix, finding eigenvectors and eigenvalues of the covariance matrix, choosing principal components to form a new feature vector, and deriving the transformed dataset.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    9 views5 pages

    Principle Component Analysi12

    Uploaded by

    neelamysr
    Principle Component Analysis (PCA) is an algorithm for dimensionality reduction and feature extraction. It transforms a dataset onto a new feature space of lower dimension while retaining as much information as possible. The steps of PCA involve getting data, mean normalization, calculating the covariance matrix, finding eigenvectors and eigenvalues of the covariance matrix, choosing principal components to form a new feature vector, and deriving the transformed dataset.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    of 5

    Principle Component Analysis (Algorithm)

     Classifier
     Dimension reduction

    Input Attributed Output Attribute


    LED Bulb Fan Motor AC Total Bill (RS)
    X Y Z T C
    1 1.5 2 5 1000

    PCA will find the principle components from dataset and sort them wrt to level of significance

    X + Y+ Z + T = C
    δ1*X + δ2*Y + δ3*Z + δ4*T = C
    0.1X +0.3Y +0.2Z+0.4T
    N

    ∑ δi=1
    i=1

    δ4*T + δ2*Y + δ3*Z =C

    Steps of PCA

    Step 1: Get some data


    Step 2: Subtract the mean /Mean Normalization (optional)
    Step 3: Calculate the covariance matrix
    Step 4: Calculate the eigenvectors and eigenvalues of the covariance matrix
    Step 5: Choosing components and forming a feature vector
    Step 6: Deriving the new data set
    y

    ∑ ( H i−H ) .(M i−M )


    COV (H,M) = i=1
    n−1

    M-
    H M H-Mean(H) Mean(M) Var (H)
    -
    23.416666
    9 39 -4.9166667 7 115.1319444 24.17361111
    -
    1.0833333 6.4166666 -
    15 56 3 7 6.951388889 1.173611111
    11.083333 30.583333
    25 93 3 3 338.9652778 122.8402778
    -
    0.0833333 1.4166666 -
    14 61 3 7 0.118055556 0.006944444
    10 50 -3.9166667 - 48.63194444 15.34027778
    12.416666
    7
    4.0833333 12.583333
    18 75 3 3 51.38194444 16.67361111
    -
    30.416666
    0 32 -13.916667 7 423.2986111 193.6736111
    2.0833333 22.583333
    16 85 3 3 47.04861111 4.340277778
    -
    20.416666
    5 42 -8.9166667 7 182.0486111 79.50694444
    5.0833333 7.5833333
    19 70 3 3 38.54861111 25.84027778
    2.0833333 3.5833333
    16 66 3 3 7.465277778 4.340277778
    6.0833333 17.583333
    20 80 3 3 106.9652778 37.00694444
    1352.416667 524.9166667

    Mean 13.9166
    (H) 7
    Mean( 62.4166
    M) 7
    COVS 122.947
    112.701
    CovP 4
    Var(H) 47.7197
    Var(H) 47.7197
    370.083
    Var(M) 3

    Cov Matrix
    COV(H,H COV(H,M
    COV (H,M) ) )
    COV(M,H COV(M,M
    ) )

    COV(X,Z
    COV(X,Y,Z) COV(X,X) COV(X,Y) )
    COV(Y,Z
    COV(Y,X) COV(Y,Y) )
    COV(Z,Z
    COV(Z.X) COV(Z,Y) )
    AX=λX
    A= Input Vector (Covariance matrix)
    X=Eigen Vector
    λ= Eigen value
    AX-λX=0
    (A-λ)X=0

    (A-λI)X=0
    let B=(A-λI)
    BX=0
    X= B−1 .0
    Determinant(B)=0

    Example
    Let A is the covariance matrix of X,Y,Z

    (A-λI)X=0
    A I
    1 2 1 1 0 0
    6 -1 0 - λ 0 1 0
    -1 -2 -1 0 0 1

    1 2 1 λ 0 0
    6 -1 0 - 0 λ 0
    -1 -2 -1 0 0 λ

    1-λ 2 1
    6 -1-λ 0
    -1 -2 -1-λ
    1 6 -1-λ - 0 1-λ 2 + -1-λ 1-λ 2
    -1 -2 -1 -2 6 -1-λ

    1(-12-(1+ λ)) - 0.(-2+2 λ+2) + (-1- λ)[(1- λ)(-1- λ)-12)]


    (-12-1- λ)+(-1- λ)[-1- λ+ λ+ λ2-12]
    (-13- λ)+(-1- λ)(-13+ λ2)
    (-13- λ)+(13- λ2+13λ- λ3)
    - λ3- λ2+12λ =0
    - λ (λ2+ λ-12)=0
    λ2+ 4λ-3λ-12 =0
    λ=-4
    λ=3
    λ =0

    clc;
    clear all;
    close all;
    A=[1 2 1; 6 -1 0; -1 -2 -1];
    [U V]=eig(A);

    clc;
    clear all;
    close all;
    H=[9 15 25 14 10 18 0 16 5 19 16 20];
    M=[39 56 93 61 50 75 32 85 42 70 66 80];
    C=cov(H,M)

    You might also like