JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.
DLV-4 Metric learning and Few shot Learning
Some of the questions have multiple answers correct.
* Indicates required question
Email
*
Your email
If a function f shown below is a valid distance metric between two vectors x and y, for a fixed square matrix M. Which of the following statements are true? [multiple options may be correct]
1 point
Eigen values of M are non-negative
f(x,y) > =0
f(x, z) <= f(x, y) + f(y, z)
Determinant of M should be zero
Find the angle in radians between the vectors v1 and v2?
1 point
Your answer
The features encoded for images of same class from a classifier preferably could have which of the following charactericstics (based on the loss function we enforce for a good classifier). [multiple options may be correct]
1 point
Their "distance" in the feature space should be low
The cosine of angle between them should be close to one
They should be independent or uncorrelated.
They should cluster together in feature space
None of the above
Suppose that we draw a shape in the plane with area 100m^2 . What is the area in m^2 after transforming the figure by the matrix.
1 point
Your answer
If we are trying to minimise a loss function shown below, what are we trying to to enforce mathematically. [multiple options may be correct]
1 point
m - Dw should be maximum
m>Dw
m<=Dw
minimise (m-Dw) till it becomes 0
Send me a copy of my responses.
Submit
Clear form
Never submit passwords through Google Forms.
reCAPTCHA
Privacy
Terms
This content is neither created nor endorsed by Google.
Report Abuse
-
Terms of Service
-
Privacy Policy
Forms
