Neural networks their strengths and weaknesses when used in experts
The latest know-how in the field of experts is the use of neural networks. This term has been borrowed from artificial intelligence systems. Technically speaking, neural networks, simply put, mimic the mechanisms of the human brain. The main characteristic of such systems is their ability to learn from their actions.
The essence and advantages of neural networks
The use of neural networks in Expert Advisors allows you to receive multiple streams of information and get one result at the output.
Before you can use a neural network for prediction, you need to train it to find and correct the patterns. The training and testing process is quite time consuming, but enables the neural network to predict the future situation based on retrospective data. When input and output data pairs occur, the neural network learns the identified dependencies and applies those dependencies to the newly received data. Thus, the network compares the obtained result with its prediction and can return to adjust the significance of certain dependencies until it gets the correct conclusions.
Two different sets of data are used to "train" a neural network: a training set and a testing set. The advantage of neural networks is that they are constantly learning by comparing their predictions to incoming data. Additionally, neural networks combine fundamental and technical data for optimal use. Networks have enough intrinsic power to identify unaccounted patterns and apply them further to predictions to achieve the most accurate output.
Disadvantages of neural networks
Unfortunately, the advantages of neural networks in forecasting in trading at the same time can be their disadvantages. The output information has the same quality as the incoming information. A neural network can detect a pattern of different types of information even in the absence of any relationship. The ability to apply intelligence without regard to emotion, a major advantage of the machine over humans, is also a disadvantage of the neural network, since with increased volatility in the marketplace, the network cannot assign weight to a sudden emotional factor.
Nowadays there are many robots whose work is based on the use of neural networks. Nevertheless, they are not the Grail that many generations of traders have been searching for. It is important to always remember the main rule of trading when using neural networks - always train the system, test it and optimize it, and success will not be long in coming.