Machine learning quiz questions TRUE or FALSE with answers, important machine learning interview questions for data science, Top 3 machine learning question set, ML exam questions
Machine Learning TRUE / FALSE Questions - SET 11
1. A multi-layer
neural network model trained using stochastic gradient descent on the same
dataset with different initializations for its parameters is guaranteed to learn
the same parameters.
(a) TRUE (b)
FALSE
View Answer
Answer: FALSE
The loss function
for a multi-layer neural network is non-convex and hence SGD is only
guaranteed to converge to a local optimum.
|
2. Stochastic
gradient descent results in a smoother convergence plot (loss vs epochs) as
compared to batch gradient descent.
(a) TRUE (b)
FALSE
View Answer
Answer: FALSE
SGD results in
noisier convergence plots compared to batch gradient descent.
|
3. The gradient of
the sigmoid with respect to an input that is very large will be infinity.
(a)
TRUE (b)
FALSE
View Answer
Answer: FALSE
Because the
sigmoid is almost at for very large values, its gradient will be almost 0.
|
**********************
No comments:
Post a Comment