Deep Learning

Deep Learning is a type of Machine Learning that processes data to detect objects, recognize conversations, translate languages, and make decisions.

The difference between Machine Learning and Deep Learning is that Machine Learning requires training data sets to learn from it while Deep Learning is more autonomous. The Machine Learning process begins with data that has been previously been processed by a person. This processing work includes data preparation, cleaning, handling missing data, feature engineering, labelling and data splitting for training and testing. In the case of Deep Learning, the model requires a person to setup the right neural layer and other parameters such as the number of hidden layers, the learning rate, the activation function, the mini-batch size and the number of epochs. Due to this difference, systems using Deep Learning will need to have very advanced GPUs (graphics processing unit) and large storage capacity.

In order to perform these analyzes, the Deep Learning system relies on its artificial neural networks. Deep Learning neural networks are a multi-layered algorithm structure. The different layers of neural networks serve as filters, going from the most general to the most subtle elements, increasing the probability of detecting and generating a correct result.

Deep Learning is based on the use of artificial neural networks. Within neural networks there are 3 types that are the most used:

• Convolutional Neural Networks, applications include computer vision and Natural Language processing (NLP)
• Recurrent Neural Networks, applications include Natural Language Processing (NLP),speech recognition, language translation
• Generative Adversarial Networks, applications include image, voice and video creation