Scribd Logo
Upload

Welcome to Scribd!

  • 4 views

    QB AMT305module 2

    Uploaded by

    Rebecca Adrian

    Copyright:

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

    QB AMT305module 2

    Uploaded by

    Rebecca Adrian
    4 views4 pages

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    4 views4 pages

    QB AMT305module 2

    Uploaded by

    Rebecca Adrian

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 4

    Question bank of module 2

    1. Suppose that you are asked to perform linear regression to learn the function that outputs
    y, given the D-dimensional input x. You are given N independent data points, and that all the
    D attributes are linearly independent. Assuming that D is around 100, would you prefer the
    closed form solution or gradient descent to estimate the regressor?

    2. Suppose you have a three class problem where class label y ∈ 0, 1, 2 and each training
    example X has 3 binary attributes X1, X2, X3 ∈0, 1. How many parameters (probability
    distribution) do you need to know to classify an example using the Naive Bayes classifier?
    3. Is principal component analysis a supervised learning problem? Justify your answer
    4. Explain feature selection and feature extraction method for dimensionality reduction
    5. Use the ID3 algorithm to construct a decision tree for the data in the following table.

    6. (a)Classifier A attains 100% accuracy on the training set and 70% accuracy on the test
    set. Classifier B attains 70% accuracy on the training set and 75% accuracy on the test
    set. Which one is a better classifier? Justify your answer.
    (b) How does bias and variance trade-off affect machine learning algorithms?
    7. Let X = R2 and C be the set of all possible rectangles in two dimensional plane
    which are axis aligned (not rotated). Show that this concept class is PAC
    learnable.
    8. The following dataset can be used to train a classifier that determines whether a
    given person is likely to own a car or not. There are three features: education level
    (primary, secondary, or university); residence (city or country); gender (female,
    male).Use ID3 Algorithm and find the best attribute at the root level of the tree 9.

    9. Consider a linear regression problem y = w1x + w0, with a training set having m
    examples (x1, y1), . . .(xm, ym). Suppose that we wish to minimize the mean 5th
    degree error (loss function) given by 1/m Σ1m(yi −w1xi − w0)5.
    1. Calculate the gradient with respect to the parameter w1.
    2. Write down pseudo-code for on-line gradient descent on w1.
    3. Give one reason in favor of on-line gradient descent compared to batch-gradient
    descent, and one reason in favor of batch over on-line.
    10. Suppose the dataset had 9700 cancer-free images from 10000 images from cancer
    patients. Find precision, recall and accuracy ? Is it a good classifier? Justify.

    11. Describe the feature selectioh method of dimensionality reduction.


    12. Consider the below frequency table which shows the frequency of each of the
    attributes. Apply NaiVe Bayes algorithm to predict the fruit type of {Yellow,
    Sweet, Long).

    13. Describe the different regression models and formulate any one error measurrements
    used in regression analysis. used in regression analysis.
    14. State ID3 algorithm, used for decision tree classification
    15. What is meant by k-fold cross validation. Given a dataset with 1600 instances, 3
    how the k-fold validation is done with k = 10.

    Use the above dataset to calculate, mean square error, mean absolute error and
    root mean square error
    16. Write a short note on logistic regression.
    17 Use the following data to construct a linear regression model for an auto
    insurance premium as a function of driving experience.

    18. Explain the procedure to reduce the dimensionality of a dataset using principal
    component analysis
    19. Given the following data on a certain set of patients seen by a doctor, can the doctor conclude
    that the person having chills, fever, mild headache and without running nose has a flu?

    You might also like