1. The positive( ) direction is the maximum directional derivative of a function.
- ANSWER gradient
2. HUAWEI HiAI built an open smart ecosystem based on the ( ), device and
chip; three-layered open architecture. It enables developers to quickly utilize
Huawei's powerful AI processing capabilities to provides a better smart
application experience for uses. - ANSWER Cloud
3. Which of the following are common application scenarios of HUAWEI
CLOUD OCR?
A. Photo recognition for cats and dogs.
B. Driving licence recognition
C. VAT invoice recognition
D. Aircraft itinerary recognition - ANSWER BCD
4. Which of the following statements are true about data conversion?
A. Image data processing, such as color space conversion, grayscale image
conversion, and geometric conversion, is a part of data conversion.
B. In machine learning, features are often normalized to ensure that the value
ranges of different variables input to the same model are the same.
C. When the data dimension is high, the existing variables need to be combined
and converted to generate new features, such as average.
D. All data needs to be converted. - ANSWER Ans: AB
5. Which of the following are common ensemble learning algorithms in
machine learning?
a. Random forest
b. GBDT
c. Polynomial regression
d. Adaboost - ANSWER ABD
,6. Which of the following statements are true about grid search based on
hyperparameter tuning?
a. Grid search exhaustively searches for all possible combinations to form a
hyperparameter value grid
b. Grid search is expensive and time-consuming
c. Grid search works well when there are relatively few hyperparameters
d. Grid search suits neural network well - ANSWER Ans: ABC
7. Which of the following products can be equipped with Ascend 310
Processors?
a. Altas 200 DK AI developer kit
b. Altas 500 AI edge computing
c. Altas 200 AI accelerator module
d. Altas 300 AI accelerator card - ANSWER ABCD
8. Which of the following are common tensor operation in MindSpore?
a. dtype()
b. astype()
c. switch()
d. abs() - ANSWER ABD
9. Which of the following development modes are supported by the ModelArts
training platform?
a. In-cloud development(CodeLab)
b. In-cloud development(Notebook+SDK)
c. On-premise development(PyCharm+PyCharm Toolkit)
d. On-premises development(IDE+SDK) - ANSWER Ans: ABCD
10. Which of the following statements are true about support vector
machines(SVMs)?
a. SVM are binary classification models. Their basic model is the linear
classifier with the largest interval defined in the feature space.
b. SVMs also have a kernel trick, which enables them to perform as a nonlinear
classifier
c. The learning algorithm of SVMs is the optimal algorithm for concave
quadratic programming
,d. Based on the structural risk minimization(SRM), SVMs build an optimal
hyperplane in the feature space so that the learner can be globally optimized. -
ANSWER ABCD
11. Which of the following statements are false about the Gated Recurrent
Unit(GRU)?
a. Unlike long short-term memory(LSTM), GRU merges the cell state and
hidden state.
b. GRU is a variant of a convolutional neural network(CNN)
c. GRU combines the forget and update gates into a single input gate
d. GRU assembles the output of the previous time step with the input of the
current time step. - ANSWER BC
12. In MindSpore, if you want to define a function for collecting information
after each step is complete, you need to inherit the ( ) class. - ANSWER
Callback
13. Which of the following statements are false about Keras?
a. Like TensorFlow, Keras is a multi-layer neural network development
package. However, Keras has simpler syntax and is easier to use.
b. Keras is a neural network development package used to build CNN
sequential models. It can not be used to build other neural network.
c. The neural network model built using Keras must be compiled before data
can be input into it for training.
d. Keras can be used as the backend of TensorFlow. - ANSWER Ans: BD
14. The average value of tensors generated using
tf.random.normal([4],2,1,tf.float32) is ( ) - ANSWER 2
15. Which of the following are part of MindSpore?
a. Mind Expression API layer
b. Mind Compiler's real-time graph-level compilation subsystem
c. CUDA and third-party chips
d. MindData - ANSWER ABD
16. Which of the following statements is false about gradient descent algorithm?
a. The gloabal gradient descent is more stable than the stochastic gradient
descent(SGD)
, b. When there are too many samples and GPUs are not used for parallel
computing, the convergence process of the global gradient descent is time-
consuming
c. When GPUs are not used for parallel computing, the mini-batch gradient
descent (MGGD) takes less time than the SGD to complete an epoch
d. Each time the global descent updates its weight, all training samples need to
be calculated. - ANSWER C
17. In the Da Vinci Architecture, which of the following computation data types
is supported by the vector unit?
a. INT16
b. BFloat16
c. FP32
d. BFloat32 - ANSWER c
18. Which of the following statements are true about the three main schools of
AI?
a. Behaviourism states that intelligence requires no knowledge, representation,
or reasoning
b. Connectionism states that AI can evolve like human intelligence
c. Behaviourism states that intelligent behaviour can only be expressed through
constant interaction between humans and the surrounding environment in the
real world.
d. Symbolism states that AI focuses on behaviour control, adaptive
computation, and evolutionary computation. - ANSWER AC
19. At a convolution layer, there are 256 5X5 convolution kernels, the size of an
input feature map is 32X32X5, the convolution step is 2, and the zero padding is
employed(padding size=1). So the size of the output feature map is ( )*( )*( ). -
ANSWER 15*15*256
20. Which of the following statements are false about the combination of the
model bias and variance?
a. A good model requires a high variance and high bias.
b. A high bias and high variance may result in overfitting
c. A high bias and low variance may result in underfitting.
d. A model with a high variance is of no application value. - ANSWER ABD
- ANSWER gradient
2. HUAWEI HiAI built an open smart ecosystem based on the ( ), device and
chip; three-layered open architecture. It enables developers to quickly utilize
Huawei's powerful AI processing capabilities to provides a better smart
application experience for uses. - ANSWER Cloud
3. Which of the following are common application scenarios of HUAWEI
CLOUD OCR?
A. Photo recognition for cats and dogs.
B. Driving licence recognition
C. VAT invoice recognition
D. Aircraft itinerary recognition - ANSWER BCD
4. Which of the following statements are true about data conversion?
A. Image data processing, such as color space conversion, grayscale image
conversion, and geometric conversion, is a part of data conversion.
B. In machine learning, features are often normalized to ensure that the value
ranges of different variables input to the same model are the same.
C. When the data dimension is high, the existing variables need to be combined
and converted to generate new features, such as average.
D. All data needs to be converted. - ANSWER Ans: AB
5. Which of the following are common ensemble learning algorithms in
machine learning?
a. Random forest
b. GBDT
c. Polynomial regression
d. Adaboost - ANSWER ABD
,6. Which of the following statements are true about grid search based on
hyperparameter tuning?
a. Grid search exhaustively searches for all possible combinations to form a
hyperparameter value grid
b. Grid search is expensive and time-consuming
c. Grid search works well when there are relatively few hyperparameters
d. Grid search suits neural network well - ANSWER Ans: ABC
7. Which of the following products can be equipped with Ascend 310
Processors?
a. Altas 200 DK AI developer kit
b. Altas 500 AI edge computing
c. Altas 200 AI accelerator module
d. Altas 300 AI accelerator card - ANSWER ABCD
8. Which of the following are common tensor operation in MindSpore?
a. dtype()
b. astype()
c. switch()
d. abs() - ANSWER ABD
9. Which of the following development modes are supported by the ModelArts
training platform?
a. In-cloud development(CodeLab)
b. In-cloud development(Notebook+SDK)
c. On-premise development(PyCharm+PyCharm Toolkit)
d. On-premises development(IDE+SDK) - ANSWER Ans: ABCD
10. Which of the following statements are true about support vector
machines(SVMs)?
a. SVM are binary classification models. Their basic model is the linear
classifier with the largest interval defined in the feature space.
b. SVMs also have a kernel trick, which enables them to perform as a nonlinear
classifier
c. The learning algorithm of SVMs is the optimal algorithm for concave
quadratic programming
,d. Based on the structural risk minimization(SRM), SVMs build an optimal
hyperplane in the feature space so that the learner can be globally optimized. -
ANSWER ABCD
11. Which of the following statements are false about the Gated Recurrent
Unit(GRU)?
a. Unlike long short-term memory(LSTM), GRU merges the cell state and
hidden state.
b. GRU is a variant of a convolutional neural network(CNN)
c. GRU combines the forget and update gates into a single input gate
d. GRU assembles the output of the previous time step with the input of the
current time step. - ANSWER BC
12. In MindSpore, if you want to define a function for collecting information
after each step is complete, you need to inherit the ( ) class. - ANSWER
Callback
13. Which of the following statements are false about Keras?
a. Like TensorFlow, Keras is a multi-layer neural network development
package. However, Keras has simpler syntax and is easier to use.
b. Keras is a neural network development package used to build CNN
sequential models. It can not be used to build other neural network.
c. The neural network model built using Keras must be compiled before data
can be input into it for training.
d. Keras can be used as the backend of TensorFlow. - ANSWER Ans: BD
14. The average value of tensors generated using
tf.random.normal([4],2,1,tf.float32) is ( ) - ANSWER 2
15. Which of the following are part of MindSpore?
a. Mind Expression API layer
b. Mind Compiler's real-time graph-level compilation subsystem
c. CUDA and third-party chips
d. MindData - ANSWER ABD
16. Which of the following statements is false about gradient descent algorithm?
a. The gloabal gradient descent is more stable than the stochastic gradient
descent(SGD)
, b. When there are too many samples and GPUs are not used for parallel
computing, the convergence process of the global gradient descent is time-
consuming
c. When GPUs are not used for parallel computing, the mini-batch gradient
descent (MGGD) takes less time than the SGD to complete an epoch
d. Each time the global descent updates its weight, all training samples need to
be calculated. - ANSWER C
17. In the Da Vinci Architecture, which of the following computation data types
is supported by the vector unit?
a. INT16
b. BFloat16
c. FP32
d. BFloat32 - ANSWER c
18. Which of the following statements are true about the three main schools of
AI?
a. Behaviourism states that intelligence requires no knowledge, representation,
or reasoning
b. Connectionism states that AI can evolve like human intelligence
c. Behaviourism states that intelligent behaviour can only be expressed through
constant interaction between humans and the surrounding environment in the
real world.
d. Symbolism states that AI focuses on behaviour control, adaptive
computation, and evolutionary computation. - ANSWER AC
19. At a convolution layer, there are 256 5X5 convolution kernels, the size of an
input feature map is 32X32X5, the convolution step is 2, and the zero padding is
employed(padding size=1). So the size of the output feature map is ( )*( )*( ). -
ANSWER 15*15*256
20. Which of the following statements are false about the combination of the
model bias and variance?
a. A good model requires a high variance and high bias.
b. A high bias and high variance may result in overfitting
c. A high bias and low variance may result in underfitting.
d. A model with a high variance is of no application value. - ANSWER ABD
