Deep Learning, which is one of the machine learning variations of today’s favorite technologies, is an increasingly developing field. Deep Learning is also a learning method, although its meaning is not understood at first. In this article, we will try to answer the question of What is Deep Learning .
What is Deep Learning?
- What is Deep Learning?
- Where is Deep Learning Used?
- How Deep Learning Works
- What is the Difference of Machine Learning with Deep Learning?
Driverless vehicles , which we often see , constitute the main technology behind the sound controllable devices. With Deep Learning, results that were previously unattainable were achieved.
Deep learning is a computer model. He can learn from pictures, text or sound. The more information he learns, the more advanced he can do. Sometimes it goes beyond human ability. Nowadays, thanks to deep learning, driverless vehicles can be developed and systems that can stop when they see the red light and automatically park when they find empty parking spaces are developed.
You can ask, of course, how Deep Learning is able to give you such great results. You need to know that there is a lot of data behind this technology. There’s nothing for free. Deep Learning’s multi-layered structure successfully accomplishes high determination of knowledge, object and concept.
The idea, which was put forward theoretically in the 1980s, is now actively put forward. There are two main answers to why it has become popular today, not in the past.
1- Deep Learning needs labeled data, also called labeled data . And that’s not exactly two or three. For example, millions of pictures and thousands of hours of video are needed to produce the system we call driverless vehicle.
2- Deep Learning Requires very high processing power. High-performance GPUs can be used to produce solutions with the power obtained. In the past, there were no high processors available today. However, thanks to the enormous processing power achieved with parallel systems, the elements that Deep Learning needs are now available.
Where is Deep Learning Used?
Driverless Vehicle : As mentioned before, driverless vehicle technologies are developed with deep learning. It functions from the interpretation of traffic lights to the monitoring of vehicles in front and behind you, and how to react to objects suddenly on the road.
Satellite and Defense : We know that satellites can see objects in the world very clearly. What if we could transfer every object on the world to this automatic system? Deep Learning seems to have done this to a large extent.
Medical research : In particular, cancer screening is intended to automatically scan for elements that could be missed by doctors’ technicians. It is aimed to detect abnormal situations with deep learning.
Industrial Automation : Thanks to Deep Learning, it is planned to reduce the damage caused by occupational safety and accidents.
In Electronics : In our management, electronic devices are aimed to be done by hearing and interpreting the speech rather than manual intervention.
How Deep Learning Works
Many deep learning methods use architectures called Neural Networks. What’s this deep? You can ask what it is.
The Neural Network consists of several layers. So we’re talking about many layers that go deep. In fact, the classic Neural Network (2-3) consists of layers. However, this number can reach up to 150 in deep learning. Although it is difficult to create in the head without going into detail, a lot of data is made up of these layers and the processing power of the GPU is processed. Thus, for example, when the parking space is empty, the driver can detect that the vehicle must park in empty space. Of course, the parking space is empty with deep learning
What is the Difference of Machine Learning with Deep Learning?
Deep Learning is a branch of machine learning. Machine learning is a broader set. In fact, the main difference is in the machine learning manual processes, while deep learning is intended to automate the entire system.