Quantum Neural Network

Many Quantum 2 proves this. Quantum mechanics and ANN 3 Figure 2: Quantum analogy of di

Social neural network algorithm such as Grover's search algorithm, Shor's factorization algorithm, etc. With this property and knowledge of classical neural network, we can create a new computational paradigm called quantum neural network (QNN). There are many ways to build QNN. Figure 2 is a quotation from [2] showing various implementation methods. In this chapter we will only consider the Menneer model and the Narayanan model.

Quantum neural network (QNN) is a neural network model based on the principle of quantum mechanics. There are two different approaches to QNN research. One is to use quantum information processing to improve the existing neural network model (and vice versa). And the other is to look for potential quantum effects in the brain. In the computational method of quantum neural network research, scientists have developed an artificial neural network model (an important task widely used in machine learning for pattern classification) to develop more efficient algorithms (for review) Attempt to combine with the advantages of. An important motivation for these studies is the difficulty in training classical neural networks, especially in big data applications. It is desirable to be able to use the characteristics of quantum computing such as quantum parallel processing and the influence of interference and entanglement as resources.

There are two main reasons to discuss quantum neural networks. People begin with the argument that the quantum process plays an important role in the living brain. For example, Roger Penrose believes that a new physics combined with quantum phenomena with general relativity can explain mental abilities such as understanding, consciousness, consciousness. However, this approach advocates studying the structure of intracellular structures such as microtubules rather than neural networks themselves. The second motivation is that classical artificial neural network domains can be extended to quantum fields by a combination of trade-offs with this field and promising new fields of quantum computing. Both of these factors show new insights into mental and brain function, as well as new and unprecedented functions in information processing. I outlined various approaches

Neural networks have multiple neural networks and their behavior is totally different, so it is difficult to describe them. But more importantly, knowing the concepts of neural networks such as quantum mechanics and calculus, and even the vast spaces away from the routine experiences of most people, it is important to know where It Is difficult. First of all, this is something they are not. The neural network is not a procedure reasoning system. It is not a decision tree or knowledge base that experts can create. Unlike expert systems with rules, neural networks do not codify explicit knowledge, no doubt, even in artificial intelligence (AI). Traditional AI reasoning is sometimes based on the "IF A THEN B" logic with a simple probability. On the other hand, neural networks use more specific stochastic tools. This is hard to understand for most people.