Please enable JavaScript.
Coggle requires JavaScript to display documents.
Introduction to Deep Learning - Coggle Diagram
Introduction to Deep Learning
Why deep learning taking off
Amount of data as help to improve deep learning algorithms performance
When the is no a large amount of data, other algorithms can have better performance than NN. Depende on the skill of feature engineering
the data and computation power have improve the perfomance of deep learning. But the improvements on deep learning algorithms allow the to need less computation power.
Idea > Code > Experiment >Idea...
The deep learning allow iterate
What is a Neural Network
ReLU (Retified linear units) _/
On some examples the y cannot be a negative value, the "line" is fixed
Supervised Learning
Image application
Convolutional NN
Sequence Data (Audio/Language)
RNN
Real State / Online add
Standard NN
Complex Application
Complex NN /Hybrid
NN Examples
Standard NN
NN Examples
Convolutional NN
Recurrent NN
Structured/Unstructured Data
Structured Data
Basically mean databases of data. Data organized.
Each of the features has a very welll defined meaning.
Unstructured Data
Raw audio, image text
deep learrning allow the computers to understand unstructured data
