LU4 Flashcards
What is machine learning
study of statistical models and algorithms that computer systems use to perform tasks without explicit instructions. Involves clustering, classification and development of predictive models
what does machine learning do for computers
it allows computers to learn without an explicit need for programming
What is deep learning
the scientific study of statistical models and algorithms that computer systems use to perform a task without explicit instructions. Neural networks like Recurrent neural networks, convolution neural networks and deep belief networks
what is deep learning modeled after
the neurons of the human brain
what is the difference between deep learning and machine learning in terms of data dependencies?
when the data is small deep learning does not perform well
what is the difference between deep learning and machine learning in terms of hardware dependencies?
deep learning requires high end machines- and a better GPU- as they do a lot of matrix multiplication operations
what is the difference between deep learning and machine learning in terms of execution time?
Typically deep learning takes more time to train, this is because of the many parameters in deep learning.
give practical examples of deep learning
medical imaging
marketing(time-series forecasting/sales)
industrial robotics
self driving cars
natural language processing
E- commerce (insights to customers based)
what is multilayer perceptron(MLP) class
A fully connected class of feedforward artificial neural networks
What does multilayer perceptron(MLP) generate
A set of outputs from a set of inputs
What does multilayer perceptron(MLP) use for training
backpropagation
How does the algorithm from MLP work?
Inputs are pushed forward through the MLP by taking the dot product of the input with the weights that exist between the input layer and the hidden layer
This dot product yields a value at the hidden layer.
Explain how multilayer perceptron(MLP) utilize activation functions
at each of their calculated layers they push the calculated output at the current layer through the functions. Once the calculated output has been pushed ,push it to the next layer by taking the dot product with the corresponding weights
At the output layer, the calculations will either be used for a backpropagation algorithm that corresponds to the activation function that was selected for the MLP (in the case of training) or a decision will be made based on the output (in the case of testing).
Name multilayer perceptron(MLP) activation functions
rectified linear units
sigmoid function
tanh
what is a Multilayer perceptron
A supervised learning algorithm that learns a function by training on a dataset, where the number of dimensions for input is the number for output
