先試後買

A. samples collected adequately cover the space of all possible scenarios

B. samples collected are skewed with each other

C. samples collected are all strongly correlated with each other

D. samples collected focus only on the most common cases

A. Machine learning is optimized to run on parallel computing and cloud computing.

B. Machine learning takes full advantage of SDKs and APIs.

C. Machine learning models take less time to train.

D. Machine learning does not require explicit coding of decision logic.

A. to smooth the gradient of the loss function in order to avoid getting trapped in small local minimas

B. to scale the gradient descent step in proportion to the gradient magnitude

C. to randomize the trajectory of the neural network parameters during training

D. to compute the gradient of the loss function with respect to the neural network parameters

A. 81 false positives and 33 false negatives

B. 73 false positives and 81 false negatives

C. 33 false positives and 81 false negatives

D. 81 false positives and 73 false negatives

A. trend, seasonality, causation

B. trend, seasonality, noise

C. trend, noise, kurtosis

D. trend, noise, covariance

A. It is rarely used today.

B. It is equal to ASCII.

C. It is encoding for Unicode characters.

D. It is encoding for Latin script.

A. hyperband

B. grid search

C. random search

D. population based training