Top 3 Machine Learning Quiz Questions with Answers explanation, Interview questions on machine learning, quiz questions for data scientist answers explained, machine learning exam questions
Machine learning MCQ - Set 11
1. Given a Neural Net with N input nodes, no hidden layers, one output node, with Entropy Loss and Sigmoid Activation Functions, which of the following algorithms (with the proper hyper-parameters and initialization) can be used to find the global optimum?
a) Stochastic Gradient Descent
b) Mini-Batch Gradient Descent
c) Batch Gradient Descent
d) All of the above
|
Answer: (d) all of the above All three can be used to find global optimum. Batch gradient descent - all the training data is taken into consideration to take a single step (one epoch).Mini-batch gradient descent - We use a batch of a fixed number of training examples which is less than the actual dataset and call it a mini-batch. (process neural net with each mini-batch is one epoch). Stochastic gradient descent – we consider 1 example at a time to take a single step. (process neural net with 1 example is 1 epoch) |
2. Say the incidence of a disease D is about 5 cases per 100 people (i.e., P(D) = 0.05). Let Boolean random variable D mean a patient “has disease D” and let Boolean random variable TP stand for "tests positive." Tests for disease D are known to be very accurate in the sense that the probability of testing positive when you have the disease is 0.99, and the probability of testing negative when you do not have the disease is 0.97. What is P(TP), the prior probability of testing positive.
a) 0.0368
b) 0.473
c) 0.078
d) None of the above
|
Answer: (c) 0.078 Given, P(D) = 0.05, P(TP | D) = 0.99, P(¬TP | ¬D) = 0.97, P(TP) = P(TP | D) P(D) + P(TP | ¬D) P(¬D) = (.99)(.05) + (1 - .97)(1 - .05) = 0.078. |
3. For a Gaussian Bayes classifier, which one of these structural assumptions is the one that most affects the trade-off between underfitting and overfitting:
a) Whether we learn the class centers by Maximum Likelihood or Gradient Descent
b) Whether we assume full class covariance matrices or diagonal class covariance matrices
c) Whether we have equal class priors or priors estimated from the data.
d) Whether we allow classes to have different mean vectors or we force them to share the same mean vector
|
Answer: (b) Whether we assume full class covariance matrices or diagonal class covariance matrices Diagonal covariance matrix satisfies the naive Bayes assumption. |
No comments:
Post a Comment