Neural networks and machine learning (Synergy Answers)

Sold 0
Refunds 0
Good feedbacks 0
Bad feedbacks 0

Ready answers to the test Neural networks and machine learning.
Educational institution: SYNERGY, MY, MIT.
Delivery: fresh.
The result of passing is 95-100 points.
Below you can find questions on the Neural Networks and Machine Learning test.

QUESTIONS FOR THE TEST:
What is the training set for perceptron training?

We launch the training vector X. In what case do the weight values not need to be changed?

We feed vector a to the perceptron input. When should weight values be increased?

We feed vector a to the perceptron input. When should weight values be reduced?

If on a given training pair the perceptron symbol does not match the desired answer, then:

How does neural network training occur?

"Learning with a teacher" is:

Synapses are called:

Dendrites are called:

Artificial Neuron

Artificial neural networks (ANNs) are machine learning models that use combinations of distributed simple operations dependent on learnable parameters to process input data. What type of ANN does not exist?

What is called an "epoch" in the perceptron learning algorithm?

The Similar Perceptron Theorem states that:

The perceptron looping theorem states that:

How many layers can a perceptron contain?

The two-layer perceptron theorem states that:

The question of choosing a step when applying the training procedure is resolved as follows:

What are Kohonen neurons called?

Under what conditions is a decision tree used in the decision making process?

What is the difference between neural network technologies and conventional expert systems?

Neural networks perform well not only in recognition, but also in image generation. But they still have problems with some things. With what exactly?

Neural networks have achieved particular success in working with images. But what can’t neural networks do?

Which type of machine learning is based on the interaction of the learning system with the environment?

When talking about neural networks and machine learning, Moore's law is often mentioned. What is its essence?

Let's say we need to calculate the necessary parameters to create the skin of an aircraft. Which area of machine learning will help us with this?

Machine learning has a number of challenges. What is the name of the one that is aimed at predicting the value of one or another continuous numerical value for the input data?

Big data is:

The most rarely used in practice are machine learning methods based on:

The k-means algorithm is designed to solve the problem:

The implementation of the supervised learning method does not require:

Convolutional neural networks are most effectively used to solve problems:

The process of training a neural network is called:

What is the input of an artificial neuron?

The strategy for avoiding local minima while maintaining stability is to

The Hopfield network is replaced by the Hamming network if:

Who created the first artificial neural network model?

When did the first mentions of artificially created humanoid creatures date back?

The modern history of artificial intelligence is associated with the emergence of learning algorithms. There are many types of them, and among them are sorting algorithms. Which one is considered the simplest?

Google's program learned to draw based on sketches made by people. What did the program take into account?

In 2016, the AlphaGo program beat one of the world chess champions, Lee Sedol. The next world championship tournament is scheduled for May 2017. Which company developed the AlphaGo AI?

What is a neuron's weight set?

What does the NET value mean?

What does the OUT value mean?

The activation function is called:

XW matrix multiplication calculates:

The activation function is used for:

The value of the activation function is:

In what case can multilayer networks not lead to an increase in computing power compared to a single layer network?

An activ
Which networks are characterized by a lack of memory?

The input layer of the network is called:

Is it possible to build a single-layer neural network with feedback?

Direct distribution networks are:

Feedback networks are:

"Learning without a teacher" is characterized by the absence of:

In which learning algorithm does the training set consist only of input vectors?

In which learning algorithm does the training set consist of both input and output vectors?

What types of neural network training do you know?

What needs to be done so that the neural network can help form a solution:

What tools are used to form a decision under conditions of uncertainty?

What tools are used to form a decision under conditions of certainty?

What tools are used to form a decision under risk conditions?

A neural network is trained if:

Network paralysis can occur when:

If the network has a very large number of neurons in hidden layers, then:

If the network contains two intermediate layers, then it models: