Scribd Logo
Upload

Welcome to Scribd!

  • Quiz Week 7 - Support Vector Machines

    Uploaded by

    charu.hitechrobot2889
    457 views3 pages
    The document contains a quiz about support vector machines (SVMs). It includes 5 multiple choice or multiple selection questions about SVMs, including: 1) When increasing or decreasing the C or sigma^2 hyperparameter would be appropriate for an SVM that is underfitting data. 2) Identifying a plot of the Gaussian kernel function with a different sigma^2 value. 3) Identifying which conditions would make the first term of the SVM objective function equal to zero. 4) Determining appropriate next steps when logistic regression underfits training data. 5) Identifying true statements about SVMs, kernels, decision boundaries, and parameter dependence on C.

    Original Description:

    Original Title

    quiz week 7 - support vector machines

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document contains a quiz about support vector machines (SVMs). It includes 5 multiple choice or multiple selection questions about SVMs, including: 1) When increasing or decreasing the C or sigma^2 hyperparameter would be appropriate for an SVM that is underfitting data. 2) Identifying a plot of the Gaussian kernel function with a different sigma^2 value. 3) Identifying which conditions would make the first term of the SVM objective function equal to zero. 4) Determining appropriate next steps when logistic regression underfits training data. 5) Identifying true statements about SVMs, kernels, decision boundaries, and parameter dependence on C.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    457 views3 pages

    Quiz Week 7 - Support Vector Machines

    Uploaded by

    charu.hitechrobot2889
    The document contains a quiz about support vector machines (SVMs). It includes 5 multiple choice or multiple selection questions about SVMs, including: 1) When increasing or decreasing the C or sigma^2 hyperparameter would be appropriate for an SVM that is underfitting data. 2) Identifying a plot of the Gaussian kernel function with a different sigma^2 value. 3) Identifying which conditions would make the first term of the SVM objective function equal to zero. 4) Determining appropriate next steps when logistic regression underfits training data. 5) Identifying true statements about SVMs, kernels, decision boundaries, and parameter dependence on C.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Flag for inappropriate content
    Download as pdf or txt
    of 3

    9/11/2017 Coursera | Online Courses From Top Universities.

    Join for Free

    Support Vector Machines


    Quiz, 5 questions

    1 1.  Suppose you have trained an SVM classi er with a Gaussian kernel, and it learned the
    point following decision boundary on the training set:

    You suspect that the SVM is under tting your dataset. Should you try increasing or decreasing
    C ? Increasing or decreasing σ 2 ?

    It would be reasonable to try increasing C . It would also be reasonable to try


    decreasing σ 2 .

    It would be reasonable to try decreasing C . It would also be reasonable to try


    decreasing σ 2 .

    It would be reasonable to try increasing C . It would also be reasonable to try


    increasing σ 2 .

    It would be reasonable to try decreasing C . It would also be reasonable to try


    increasing σ 2 .

    1 2.  The formula for the Gaussian kernel is given by similarity(x, l(1) ) = exp (−
    ||x−l
    (1)

    2
    2
    ||
    ) .
    point
    2σ

    The gure below shows a plot of f 1 = similarity(x, l


    (1)
    ) when σ 2 .
    = 1

    Which of the following is a plot of f 1 when σ 2 = 0.25 ?

    https://www.coursera.org/learn/machine-learning/exam/PLdSa/support-vector-machines 1/3
    9/11/2017 Coursera | Online Courses From Top Universities. Join for Free

    Support Vector Machines


    Quiz, 5 questions

    Figure 4.

    Figure 3.

    Figure 2.

    https://www.coursera.org/learn/machine-learning/exam/PLdSa/support-vector-machines 2/3
    9/11/2017 Coursera | Online Courses From Top Universities. Join for Free

    1 3.  The SVM solves


    Support
    point
    Vector Machines
    Quiz, 5 questions m (i) T (i) (i) T (i) n 2
    minθ  C ∑ y cost1 (θ x ) + (1 − y )cost0 (θ x )+∑ θj
    i=1 j=1

    where the functions cost0 (z) and cost1 (z) look like this:

    The rst term in the objective is:

    m (i) T (i) (i) T (i)


    C∑ y cost1 (θ x ) + (1 − y )cost0 (θ x ).
    i=1

    This rst term will be zero if two of the following four conditions hold true. Which are the two
    conditions that would guarantee that this term equals zero?

     For every example with y (i) = 0 , we have that θT x(i) ≤ 0 .

    For every example with y (i) = 0 , we have that θT x(i) ≤ −1 .

    For every example with y (i) = 1 , we have that θT x(i) ≥ 0 .

    For every example with y (i) = 1 , we have that θ T


    x
    (i)
    ≥ 1 .

    1 4.  Suppose you have a dataset with n = 10 features and m = 5000 examples.
    point
    After training your logistic regression classi er with gradient descent, you nd that it has
    under t the training set and does not achieve the desired performance on the training or
    cross validation sets.

    Which of the following might be promising steps to take? Check all that apply.

    Increase the regularization parameter λ.

    Create / add new polynomial features.

    Use an SVM with a linear kernel, without introducing new features.

    Use an SVM with a Gaussian Kernel.

    1 5.  Which of the following statements are true? Check all that apply.
    point
    If you are training multi-class SVMs with the one-vs-all method, it is

    not possible to use a kernel.

    Suppose you have 2D input examples (ie, x(i) ∈ R


    2
    ). The decision boundary of the
    SVM (with the linear kernel) is a straight line.

    The maximum value of the Gaussian kernel (i.e., sim(x, l(1) )) is 1.

    If the data are linearly separable, an SVM using a linear kernel will

    return the same parameters θ regardless of the chosen value of

    C (i.e., the resulting value of θ does not depend on C ).

    I, Mark R. Lytell, understand that submitting work that isn’t my own may result in permanent failure of this course
    or deactivation of my Coursera account. Learn more about Coursera’s Honor Code

    Submit Quiz

      

    https://www.coursera.org/learn/machine-learning/exam/PLdSa/support-vector-machines 3/3

    You might also like