Brief explanation by Nadia Zaifulizan
Deep Learning is about neural networks. The structure of neural networks consist of nodes (neurons) and connect together.
They are used to receive inputs, perform complex calculations or operations, and produce outputs.
Examples of its use are Classification. Each ‘Classifier’ would receive data input or information, process it, and produce a prediction or score, to ‘classify’ if the data is nearer to the ‘confidence’ level of a specific criteria, or if it is not (means it is the opposite or furthest from the particular criteria).
Because the predictive value needs to be improved, the neural network needs ‘Training’. This helps the neural network makes predictions as accurate as possible, close to the actual output. By running many training examples to reduce the difference or ‘Cost’, which is the difference between the current output and the output that is known to be correct, the accuracy of the neural network is eventually improved.