AI will be the biggest opportunity in the technology industry(electronics manufacturing expo news)

01 Aug 2019

Looking back over the past 250 years, from the beginning of the industrial revolution, we emerged from the industrial revolution in the UK. The invention of the steam engine enhanced the ability of humans to use kinetic energy. A electronics manufacturing expo steam engine replaced one hundred laborers. Most of the world population was still on the farm. Manual labor, such manual labor has been replaced by steam engines, just as the United States has been highly automated, from the return to the industrial revolution, the steam engine was also invented, largely replacing many artificial labor, the industrial revolution also brought population migration, From the rural to the urban areas, the industrial revolution has had a profound impact on society. Over the past 200 years, we can see that such industrial electronics manufacturing expo revolutions, the use of science and technology, and the emergence of various technologies have profoundly affected all aspects of our lives.

Of course, there are also many flaws. For decades, for example, London during the Industrial Revolution, there was a lot of smog and smoke, which was caused by the use of coal-driven electronics manufacturing expo steam engines, and the coal workers working in the coal plants also suffered from breathing. The torment of disease is a challenge that is brought about by the convenience of technology. How to carry out air treatment and reduce respiratory diseases is a big challenge.

At the same time of technological development, we also need to deal with the consequences of technology. Nowadays, the technical artificial intelligence we are facing is no exception. I have heard a lot of hot technical words that I have heard. I will give you a brief introduction. Artificial intelligence, the term was born in 1956. The goal is to imitate artificial intelligence on the machine, which is a very big electronics manufacturing expo goal. We are still far from achieving such a goal. Within the scope of artificial intelligence, one of the sub-categories is growing rapidly. The next speaker will give you a detailed explanation of machine learning. Machine learning is another way to develop artificial intelligence. The past is Programming, you can program that you already have the knowledge of such a domain to solve this electronics manufacturing expo problem, and the form of solving the problem is through programming, so you can already be an expert in this field.

The machine learning path is different. We collect data through a large amount, learn through machines, and learn by data structure, such as learning some image objects, languages, word sequences, etc. In the field of machine learning, there is another method. Learning, that is, a specific algorithm, is a direction inspired by the brain. Our brain is a very complicated device, collecting information, hundreds of billions of brain electronics manufacturing expo neurons, transmitting information, for example, so many people present, through thousands The brain of the billion brain neurons performs information processing and then transmits the information to everyone present. We still don't understand the inner workings of the brain. I will explain this piece to you later, but I want to say that the inspiration for deep learning comes from the brain's working mechanism.

The trade-off between learning and programming, the red line is the programming line. If you want to hire a programmer to solve the problem for you, this cost is increasing year by year. The cost is increasing year by year from 1980 to 2040. The programmer is very expensive, and what is needed is Expert programmer. So relatively speaking, from the beginning of development, our machine learning, since the 1980s, the electronics manufacturing expo cost is very high, the computer is very expensive, but now the cost of artificial intelligence and machine learning is constantly going down, the machine learning electronics manufacturing expo algorithm and programming cost in 2012 is roughly The same, from 2012 onwards, we are declining year by year. We use different data sets. We don't need to be fully familiar with the knowledge of this field. We don't need to fully understand this object, but if we get a lot of instance data, our machine can continue through specific learning algorithms. To solve the problem, we used the neural network processing system to propose such a theory 30 years ago, which is the neural network model.

The brain neural network can effectively handle complex data sets, and can also handle hundreds of millions of image data. I believe that everyone here is very familiar with the Alphago War. In 2017, he defeated the Ke Jie World Go Championship, and was shocked by the time. Just because of the electronics manufacturing expo complexity of the World Go Championship, humans still have a great resonance, thinking that the machine defeats human beings and transcends humanity.

Before the Alphago defeated the world champion, he thought that the machine could learn well, but still could not beat humanity. We used Ke Jie’s words to talk to the Alphago last year and felt that its way of playing chess is very close to human beings. Today, we are playing chess like the god of Go. It is very innovative in every step of chess, and it has not appeared before. Very innovative, if it is human, we call it a go genius. Because it is through artificial intelligence, the Alphago has an innovative electronics manufacturing expo chess game, so you can also see that through the AI ​​machine learning, the machine can also achieve super-goal innovation.

A very simple version of the human brain electronics manufacturing expo mechanism, there are many neurons in the brain, about hundreds of billions, they work in parallel, which is different from digital computers, because computers have processors, have memory, and Both hardware have bottlenecks. Memory is actually a connection between neurons, that is, there is a synapse. In fact, ten tenth-order neurons have ten connections, and the number of synapses in the brain is very large, that is, the brain. The memory is huge. Can store a lot of data, this is the benefit of the brain, the problem is that the human brain is a millisecond transmission speed, so it is much slower than the computer, because in nature, I don't have so many processing electronics manufacturing expo layers, and the computer is faster. In fact, the world we live in is a world of milliseconds. It doesn't need such a fast speed. The information transfer in our brain is a very complicated process, as well as the storage, processing, etc. of information, let us solve a lot of problems. This is very crucial.

There are also some challenges, such as having many neurons, synapses, etc., and ultimately around the central nervous system, which is the most important organ. We need to know if the change of a connection will change the input and output of the information. If this connection is changed, it will be affected. At that time, Jeffrey and I got the neural algorithm, that is, how to make the synapse Dealing with correct neurotransmitters, almost all scientists and engineers felt that this was impossible before. I want to tell some young people, don't trust the experts, because many electronics manufacturing expo experts always know that this can't be done, but he doesn't Know how this can be done. So we invented the Boltzmann machine and are still using it in network neural learning.

Leading Exhibitors

electronic equipment China
Recommended Exhibition
China PCB show
SMT Exhibition
Join and Follow Us
Exhibition Hours
Day Time
Wednesday, August 28 10:00-17:00
Thursday, August 29 09:30-17:00
Friday,August 30 09:30-16:30
Visitor Hotline
International Callers:86-21-2231-7105
From within China:400-650-5611

We use cookies to operate this website and to improve its usability. Full details of what cookies are, why we use them and how you can manage them can be found by reading our Privacy & Cookies page. Please note that by using this site you are consenting to the use of cookies.