This article lets you know what Deep Learning is. And not from a tech perspective. We aim to let you know how you can use the technology to make you better, or how you can use it to make things that were not otherwise possible.
Deep Learning is by far the hottest in artificial intelligence.
What makes Deep Learning unique is that you train the computer system. Usually, you do IT development by coding a program. It does not happen in Deep Learning.
In Deep Learning, you have several sub-components that look for patterns that can explain a given behaviour. Sometimes some component is of great importance to explaining the behaviour. Other times it has no or little effect on the behaviour you are looking at. Conceptually, therefore, one often can look at a Deep Learning system in the same way as a brain that processes information.
What is smart about processing data in this way is that the patterns that can best describe the behaviour emerge one in a learning process.
For example, one thinks of a Deep Learning algorithm in understanding human handwriting. By showing such a system may be 1000 ways to write the number “3” or the word “dog” with different people’s handwriting, the system learns which patterns characterize whether a number is “3” or what determines how the number “10 ” write in hand.
Some compare what happens in Deep Learning to human intuition. If you ask a person why they can read the number “3” almost no matter how it is written, then few people can say why it is possible for them. But we have no doubt that those we see are the number “3” and not “6”.
Deep Learning algorithms work the same way. There are no rules in the system that determine what defined the number three written as handwriting. But based on the knowledge it has about 1000 ways in which the number three can be written and perhaps 10,000 ways in which the number cannot be written, the system is aided to read with even the most sloppy handwritten numbers and interpret it correctly.
Another great strength of Deep Learning is that it is a technology that can be much closer to our everyday lives than conventional IT. Standard IT systems are oriented around math. In order to work, they assume that the environment they are looking at can be described by mathematical formulas. It is useful if you work with spreadsheets or design a banking system. But if you want to make services that interact with people on our terms. That is, speech, text and visual – then the Deep Learning Technologist is superior.
Deep Learning can be done so advanced it could make predictions about human intentions and rational behaviour. In the example of the curtain truck, this could mean that the system could with a certain probability predict the type of curtains the customer would end up buying.
This would be based on the knowledge of when the day, the way, the week the call was made, how far the conversation was. How many clarifying questions the customer asked. When in the conversation, the customer inquired about the price, etc.
The predictions can be made by looking at patterns from previous customers’ buying behavior which makes it possible to make it probable which product a new customer would prefer. This opens up a whole new level of customer service.
It is because of examples like this that some conceptual describes Deep Learning as systems with intuition. The system cannot explain why it thinks a given customer will be interested in a given product. However, if the system has enough data to learn, it will be able to predict a given customer’s behaviour with good accuracy.