Toán tin vuotlen.com

Free Code Camp Machine Learning With Python

Introduction: Machine Learning Fundamentals

Which statement below is false?

Neural networks are modeled after the way the human brain works.

Computer programs that play tic-tac-toe or chess against human players are examples of simple artificial intelligence.

Machine learning is a subset of artificial intelligence.

Introduction to TensorFlow

Which of the following is not a type of tensor?
Variable
Flowing
Placeholder
SparseTensor
Constant

Core Learning Algorithms

Which type of analysis would be best suited for the following problem?:

You have the average temperature in the month of March for the last 100 years. Using this data, you want to predict the average temperature in the month of March 5 years from now.

Multiple regression
Correlation
Decision tree
Linear regression

Core Learning Algorithms: Working with Data

What does the pandas .head() function do?
Returns the number of entries in a data frame.
Returns the number of columns in a data frame.
By default, shows the first five rows or entries in a data frame.

Core Learning Algorithms: Training and Testing Data

What is categorical data?
Another term for one-hot encoding.
Any data that is not numeric.
Any data that is represented numerically.

Core Learning Algorithms: The Training Process

What are epochs?
The number of times the model will see the same data.
A type of graph.
The number of elements you feed to the model at once.

Core Learning Algorithms: Classification

What is classification?
The process of separating data points into different classes.

Predicting a numeric value or forecast based on independent and dependent variables.
None of the above.

Core Learning Algorithms: Building the Model

What kind of estimator/model does TensorFlow recommend using for classification?

LinearClassifier
DNNClassifier
BoostedTreesClassifier

Core Learning Algorithms: Clustering

Which of the following steps is not part of the K-Means algorithm?
Randomly pick K points to place K centeroids.
Assign each K point to the closest K centeroid.
Move each K centeroid into the middle of all of their data points.
Shuffle the K points so they're redistributed randomly.
Reassign each K point to the closest K centeroid.

Core Learning Algorithms: Hidden Markov Models

What makes a Hidden Markov model different than linear regression or classification?

It uses probability distributions to predict future events or states.
It analyzes the relationship between independent and dependent variables to make predictions
It separates data points into separate categories.

Core Learning Algorithms: Using Probabilities to make Predictions

What TensorFlow module should you import to implement .HiddenMarkovModel()?
tensorflow.keras
tensorflow_gpu
tensorflow_probability

Neural Networks with TensorFlow

A densely connected neural network is one in which...:

all the neurons in the current layer are connected to one neuron in the previous layer.
all the neurons in each layer are connected randomly.
all the neurons in the current layer are connected to every neuron in the previous layer.

Neural Networks: Activation Functions

Which activation function squishes values between -1 and 1?
ReLU (Rectified Linear Unit)
Tanh (Hyperbolic Tangent)

Sigmoid

Neural Networks: Optimizers

What is an optimizer function?

A function that increases the accuracy of a model's predictions.
A function that implements the gradient descent and backpropagation algorithms for you.
A function that reduces the time a model needs to train.

Neural Networks: Creating a Model

Fill in the blanks below to build a sequential model of dense layers:

model = __A__.__B__([

    __A__.layers.Flatten(input_shape=(28, 28)),

    __A__.layers.__C__(128, activation='relu'),

    __A__.layers.__C__(10, activation='softmax')

])


A: keras

B: Sequential

C: Dense

 

A: tf

B: Sequential

C: Categorical

 

A: keras

B: sequential

C: dense

Convolutional Neural Networks

Dense neural networks analyze input on a global scale and recognize patterns in specific areas. Convolutional neural networks...:

also analyze input globally and extract features from specific areas.
do not work well for image classification or object detection.

scan through the entire input a little at a time and learn local patterns.

Convolutional Neural Networks: The Convolutional Layer

What are the three main properties of each convolutional layer?

Input size, the number of filters, and the sample size of the filters.
Input size, input dimensions, and the color values of the input.
Input size, input padding, and stride.

Creating a Convolutional Neural Network

Fill in the blanks below to complete the architecture for a convolutional neural network:

model = models.__A__()
model.add(layers.__B__(32, (3, 3), activation='relu', input_shape=(32, 32, 3)))
model.add(layers.__C__(2, 2))
model.add(layers.__B__(64, (3, 3), activation='relu'))
model.add(layers.__C__(2, 2))
model.add(layers.__B__(32, (3, 3), activation='relu'))
model.add(layers.__C__(2, 2))

 

A: Sequential

B: add

C: Wrapper

 

A: keras

B: Cropping2D

C: AlphaDropout

 

A: Sequential

B: Conv2D

C: MaxPooling2D

Convolutional Neural Networks: Evaluating the Model

What is not a good way to increase the accuracy of a convolutional neural network?
Augmenting the data you already have.
Using a pre-trained model.
Using your test data to retrain the model.

Convolutional Neural Networks: Picking a Pretrained Model

Fill in the blanks below to use Google's pre-trained MobileNet V2 model as a base for a convolutional neural network:

base_model = tf.__A__.applications.__B__(input_shape=(160, 160, 3),
                                               include_top=__C__,
                                               weights='imagenet'
                                               )

A: keras

B: MobileNetV2

C: False


A: Keras

B: MobileNetV2

C: True


A: keras

B: mobile_net_v2

C: False

Natural Language Processing With RNNs

Natural Language Processing is a branch of artificial intelligence that...:
deals with how computers understand and process natural/human languages.
translates image data into natural/human languages.
is focused on translating computer languages into natural/human languages.

 

Natural Language Processing With RNNs: Part 2

Word embeddings are...:

an unordered group of encoded words that describes the frequency of words in a given document.
a group of encoded words that preserves the original order of the words in a given document.
a vectorized representation of words in a given document that places words with similar meanings near each other.

Natural Language Processing With RNNs: Recurring Neural Networks

What is true about Recurrent Neural Networks?

1: They are a type of feed-forward neural network.
2: They maintain an internal memory/state of the input that was already processed.
3: RNN's contain a loop and process one piece of input at a time.
4: Both 2 and 3.

Natural Language Processing With RNNs: Sentiment Analysis

Fill in the blanks below to create the model for the RNN:

model = __A__.keras.Sequential([
    __A__.keras.layers.__B__(88584, 32),
    __A__.keras.layers.__C__(32),
    __A__.keras.layers.Dense(1, activation='sigmoid')
])

A: tensor_flow

B: embedding

C: LSTM


A: tf

B: Embedding

C: AlphaDropout


A: tf

B: Embedding

C: LSTM

Natural Language Processing With RNNs: Making Predictions

Before you make a prediction with your own review, you should...:
decode the training dataset and compare the results to the test data.
use the encodings from the training dataset to encode your review.
assign random values between 0 and the maximum number of vocabulary in your dataset to each word in your review.

Natural Language Processing With RNNs: Create a Play Generator

Fill in the blanks below to create the training examples for the RNN:

char_dataset = tf.data.__A__.__B__(text_as_int)

A: Dataset

B: from_tensor_slices


A: data

B: from_tensors


A: DataSet

B: from_generator

Natural Language Processing With RNNs: Building the Model

Fill in the blanks below to complete the build_model function:

def build_mode(vocab_size, embedding_dim, rnn_units, batch_size):
    model = tf.keras.Sequential([
        tf.keras.layers.Embedding(vocab_size,
                                  embedding_dim,
                                  batch_input_shape=[batch_size, None]),
        tf.keras.layers.__A__(rnn_units,
                              return_sequences=__B__,
                              recurrent_initializer='glorot_uniform),
        tf.keras.layers.Dense(__C__)
    ])
    __D__

A: ELU

B: True

C: vocab_size

D: return model


A: LSTM

B: False

C: batch_size

D: return model


A: LSTM

B: True

C: vocab_size

D: return model

Natural Language Processing With RNNs: Training the Model

Fill in the blanks below to save your model's checkpoints in the ./checkpoints directory and call the latest checkpoint for training:

checkpoint_dir = __A__
checkpoint_prefix = os.path.join(checkpoint_dir, 'ckpt_{epoch}')

 
checkpoint_callback = tf.keras.callbacks.__B__(
    filepath=checkpoint_prefix,
    save_weights_only=True
)

 
history = model.fit(data, epochs=2, callbacks=[__C__])


A: './training_checkpoints'

B: ModelCheckpoint

C: checkpoint_prefix

 

A: './checkpoints'

B: ModelCheckpoint

C: checkpoint_callback


A: './checkpoints'

B: BaseLogger

C: checkpoint_callback

Reinforcement Learning With Q-Learning

The key components of reinforcement learning are...
environment, representative, state, reaction, and reward.
environment, agent, state, action, and reward.
habitat, agent, state, action, and punishment.

Reinforcement Learning With Q-Learning: Part 2

What can happen if the agent does not have a good balance of taking random actions and using learned actions?

The agent will always try to minimize its reward for the current state/action, leading to local minima.

The agent will always try to maximize its reward for the current state/action, leading to local maxima.

Reinforcement Learning With Q-Learning: Example

Fill in the blanks to complete the following Q-Learning equation:

Q[__A__, __B__] = Q[__A__, __B__] + LEARNING_RATE * (reward + GAMMA * np.max(Q[__C__, :]) - Q[__A__, __B__])

A: state

B: action

C: next_state


A: state

B: action

C: prev_state


A: state

B: reaction

C: next_state

Conclusion

Most people that are experts in AI or machine learning usually...:
have one specialization.
have many specializations.
have a deep understanding of many different frameworks.