NEW
Font size
WorksheetsANN Gr12
Total questions: 13
Worksheet time: 7mins
Every Artificial Neural must have at least ______
ONe layer
Twi layers
Three layers
Four layers
is a series of algorithms that aims to understand
the relationships in a set of data using a
process inspired by the human brain
Cryptography
K-nearest value
Artificial neural network
A mathematical function that determines the output of a node, given the inputs.
Sigmoid function
Transfer function
Activation function
This entire journey of the signals, starting at the input layer, going through the hidden
layers and coming out through the output layer once.
Session
Epoch
Step
Whenever you want to do mathematical computations with arrays, you need to
import this module/library.
Numpy
Math
Lib
The attributes of the class neuron in python :
inputs , outputs
weights, bias
weights,inputs
when you want to create a numpy array, you have to call numpy‟s array
function like this:
X=numpy.array([1,3])
X=numpy[1,3]
X=array([1,3])
A network of Neurons in which input moves forward, starting at the input layer, going
through the hidden layers and coming out through the output layers.
Backward propagation
Feedforward
current neural network
Using the error to modify the weight of the connection between the layers called ...........
Forward propagation
Backward propagation
This is a measure of how fast the training needs to get completed. A larger training rate
means that the steps towards the minimum error will be large,
Learning rate
Optimisation
Calculate weight
Some of the Python libraries available that do this whole work of
training, error calculation and weight adjustment
Scikit
tensorflow
Math
Lib
predict library
pip install -U scikit-learn
The line of code is used to:
Install the Scikit-learn package
Install face recognition files
Install the library to do math claculations
Using the MLPClassifier that we don‟t need to initialize the weights and biases. These are
automatically and randomly initialized by the network.
True
False
