Access Millions of academic & study documents

Neural Network

Content type
User Generated
Subject
Data Analytics
School
University of San Francisco
Type
Homework
Showing Page:
1/3
Neural Network
Fig 1. Test Loss = 0.043 < 1
The activity of neural networks is comparable to that of the human brain (Smilkov and
Carter, n.d.). The discipline of AI helps computer systems to identify patterns and tackle
common problems. It's a technique for creating computer software that learns from the data
input. It's loosely based on our understanding of how the human brain functions. But how do
neural networks operate in practice? To begin with the explanation, a neural network is a
collection of software "neurons" that are constructed and linked together, enabling them to
communicate with one another (IBM Cloud Education, 2020). The network is then asked to
solve a problem, which it tries over and over again, each time strengthening the connections that
lead to success and weakening the connections that lead to a problem or even failure.
In the neural network simulator, I got 0.043 test loss and 0.020 training loss for my setup.
To achieve this, I used an engineering feature that included adding x squared and sine x features
to a total of six input features and setting the learning rate to 0.03. The learning was unstable

Sign up to view the full document!

lock_open Sign Up
Showing Page:
2/3
under these circumstances and needed to be modified to function better. I didn't further change
the Tanh to the ReLU activation function. Because the test loss was so significant, I additionally
applied L2 regularization to minimize overfitting. With eight neurons, the performance was
greater with one hidden layer than with the other layers I tried. The neural network is learning
the data, as we can see in the simulation. The configuration I used seemed to be correctly
identifying the spiral data.

Sign up to view the full document!

lock_open Sign Up
Showing Page:
3/3

Sign up to view the full document!

lock_open Sign Up
Unformatted Attachment Preview
Neural Network Fig 1. Test Loss = 0.043 < 1 The activity of neural networks is comparable to that of the human brain (Smilkov and Carter, n.d.). The discipline of AI helps computer systems to identify patterns and tackle common problems. It's a technique for creating computer software that learns from the data input. It's loosely based on our understanding of how the human brain functions. But how do neural networks operate in practice? To begin with the explanation, a neural network is a collection of software "neurons" that are constructed and linked together, enabling them to communicate with one another (IBM Cloud Education, 2020). The network is then asked to solve a problem, which it tries over and over again, each time strengthening the connections that lead to success and weakening the connections that lead to a problem or even failure. In the neural network simulator, I got 0.0 ...
Purchase document to see full attachment
User generated content is uploaded by users for the purposes of learning and should be used following Studypool's honor code & terms of service.
Studypool
4.7
Indeed
4.5
Sitejabber
4.4

Similar Documents