Scribd Logo
Upload

Welcome to Scribd!

  • 7 views

    Assignment 5 Solution

    Uploaded by

    Allamraju108
    Useful

    Copyright:

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

    Assignment 5 Solution

    Uploaded by

    Allamraju108
    7 views6 pages
    Useful

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Useful

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    7 views6 pages

    Assignment 5 Solution

    Uploaded by

    Allamraju108
    Useful

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 6

    Course -Introduction to Machine Learning

    Assignment- Week 5 (Logistic Regression, SVM, Kernel Function, Kernel


    SVM)
    TYPE OF QUESTION: MCQ/MSQ
    Number of Question: 10 Total Marks:10x2 =20
    __________________________________________________________________
    Question 1:
    What would be the ideal complexity of the curve which can be used for separating the
    two classes shown in the image below?

    A) Linear
    B) Quadratic
    C) Cubic
    D) insufficient data to draw conclusion

    Correct Answer: A
    Detailed Solution: The blue point in the red region is an outlier. The rest of the data is
    linearly separable.
    __________________________________________________________________

    Question 2:

    Suppose you have a dataset with n=10 features and m=1000 examples. After training a
    logistic regression classifier with gradient descent, you find that it has high training error
    and does not achieve the desired performance on training and validation sets. Which of
    the following might be promising steps to take?
    1. Use SVM with a non-linear kernel function
    2. Reduce the number of training examples
    3. Create or add new polynomial features

    A) 1, 2
    B) 1, 3
    C) 1, 2, 3
    D) None
    Correct Answer: B
    Detailed Solution: As logistic regression did not perform well, it is highly likely that the
    dataset is not linearly separable. SVM with a non-linear kernel works well for
    non-linearly separable datasets. Creating new polynomial features will also help in
    capturing the non-linearity in the dataset.
    __________________________________________________________________

    Question 3:

    In logistic regression, we learn the conditional distribution p(y|x), where y is the class
    label and x is a data point. If h(x) is the output of the logistic regression classifier for an
    input x, then p(y|x) equals:

    𝑦 (1−𝑦)
    A. ℎ(𝑥) (1 − ℎ(𝑥))
    𝑦 (1−𝑦)
    B. ℎ(𝑥) (1 + ℎ(𝑥))
    1−𝑦 𝑦
    C. ℎ(𝑥) (1 − ℎ(𝑥))
    𝑦 (1+𝑦)
    D. ℎ(𝑥) (1 + ℎ(𝑥))

    Correct Answer: A
    Detailed Solution: Refer to the lecture.
    __________________________________________________________________

    Question 4:

    The output of binary class logistic regression lies in the range:


    A. [-1,0]
    B. [0,1]
    C. [-1,-2]
    D. [1,10]

    Correct Answer: B
    Detailed Solution: The output of binary class logistic regression lies in the range:
    [0,1].

    __________________________________________________________________

    Question 5:

    State whether True or False.


    “After training an SVM, we can discard all examples which are not support
    vectors and can still classify new examples.”
    A) TRUE
    B) FALSE

    Correct Answer: A
    Detailed Solution : Using only the support vector points, it is possible to classify new
    examples.
    __________________________________________________________________

    Question 6:

    Suppose you are dealing with a 3-class classification problem and you want to train a
    SVM model on the data. For that you are using the One-vs-all method. How many
    times do we need to train our SVM model in such a case?
    A) 1
    B) 2
    C) 3
    D) 4

    Correct Answer: C
    Detailed Solution: In a N-class classification problem, we have to train the SVM N
    times in the one vs all method.
    __________________________________________________________________
    __________________________________________________________________

    Question 7:

    What is/are true about kernels in SVM?

    1. Kernel function can map low dimensional data to high dimensional space
    2. It’s a similarity function

    A) 1
    B) 2
    C) 1 and 2
    D) None of these.

    Correct Answer: C
    Detailed Solution: Kernels are used in SVMs to map low dimensional data into high
    dimensional feature space to classify non-linearly separable data. It also acts as a
    similarity function.
    _________________________________________________________________

    Question 8:

    If g(z) is the sigmoid function, then its derivative with respect to z may be written in
    term of g(z) as

    A) g(z)(g(z)-1)
    B) g(z)(1+g(z))
    C) -g(z)(1+g(z))
    D) g(z)(1-g(z))

    Correct Answer: D
    Detailed Answer:
    −𝑧
    𝑑 1 𝑒 1 1
    𝑔'(𝑧) = 𝑑𝑧
    ( −𝑧 ) = −𝑧 2 = −𝑧 (1 − −𝑧 ) = 𝑔(𝑧)(1 − 𝑔(𝑧))
    1+𝑒 (1+𝑒 ) 1+𝑒 1+𝑒
    __________________________________________________________________
    Question 9:

    Below are the labelled instances of 2 classes and hand drawn decision boundaries for
    logistic regression. Which of the following figures demonstrates overfitting of the
    training data?

    A) A
    B) B
    C) C
    D) None of these

    Correct Answer: C
    Detailed Solution: In figure 3, the decision boundary is very complex and unlikely to
    generalize the data.
    __________________________________________________________________
    Question 10:

    What do you conclude after seeing the visualization in the previous question (Question
    9)?

    C1. The training error in the first plot is higher as compared to the second and third
    plot.
    C2. The best model for this regression problem is the last (third) plot because it
    has minimum training error (zero).
    C3. Out of the 3 models, the second model is expected to perform best on
    unseen data.
    C4. All will perform similarly because we have not seen the test data.

    A) C1 and C2
    B) C1 and C3
    C) C2 and C3
    D) C4

    Correct Answer: B
    Detailed Solution: From the visualization, it is clear that the misclassified samples
    are more in the plot A when compared to B and C. So, C1 is correct. In figure 3, the
    training error is less due to complex boundaries. So, it is unlikely to generalize the
    data well. Therefore, option C2 is wrong.
    The first model is very simple and underfits the training data. The third model is very
    complex and overfits the training data. The second model compared to these models
    has less training error and is likely to perform well on unseen data. So, C3 is correct.
    We can estimate the performance of the model on unseen data by observing the
    nature of the decision boundary. Therefore, C4 is incorrect.

    __________________________________________________________________
    End

    You might also like