| The Human Brain The human brain is a unique and most amazing organ. It gives us the power to think, plan, speak, imagine, conceive and transform ideas into tangible things.
Photo of A Human Brain
(Credit: University of Nottingham, School of Mathematical Sciences)
Einstein’s brain - found to have more glial cells than others The human brain, spinal cord and peripheral nerves make up a complex, integrated information-processing and control system. The scientific study of the brain and nervous system is called neuroscience or neurobiology. Although the brain accounts for less than 2% of a person's weight, it consumes 20% of the body's energy, estimated to be about 20 Watts. The brain also uses up 20% of total body oxygen consumption. The human brain contains roughly 100 billion neurons, and each neuron is linked to other neurons with up to 10,000 synaptic connections 
PET-Scan Image: Showing energy consumption of human brain 
Drawing depicting interconnection between two neurons in brain
(Credit: Neurevolution.net) 
A neural network in the brain demonstrating the complexity of neural connections
(Credit: Williamette University, Oregon) Neural Trading Technology 
Neural Trading Technology is one of the most advanced automated FOREX systems being developed. Trading systems based on neural networks is capable of adapting to changing market conditions as compared to traditional systems with fixed rules. Darwin’s Theory of Natural Selection is often misunderstood to be “the fittest or the strongest survive”. If fact, what the Darwinian Theory of Evolution stipulates is that “the most adaptive survive”. Likewise, a sound trading system, be it manual or automated must also be able to evolve and adapt to the changing market conditions. Only adaptive trading systems can survive and thrive in a highly volatile and ever-changing market conditions. How Neural Trading Technology Works Neural Network technology is a form of artificial intelligence that mimics how the brain works. A Neural Network comprises many processing components or nodes that work in parallel. The nodes are interconnected with each other in each layer, and the layers are also interconnected. These multi-layers of interconnected nodes form the information processing system that approximates complex mathematical functions for solving specific problems. The number of nodes, and hence the network size, increases as the complexity of the task increases. Training a Neural Network Neural Networks, like their biological counterparts learn by example much akin to how we learn to shoot a basketball, or return a fast coming top-spin tennis ball. With each try, we learn to shoot the basketball or return the tennis ball better. A neural network can adapt and learn to perform many different functions such as pattern recognition, image processing and trend analysis. The learning process takes place as the neural network developer feeds the network of interconnected nodes a variety of real-life examples, called training sets. During training or learning, the neural network creates connections and learns patterns based on the input and output data sets. Each pattern creates a unique configuration of network structure with a unique set of connection weights. Each connection weight builds on previous decision nodes, propagating down to a final output or decision. During each of the training cycles, the neural network adapts to changing inputs and learns patterns or trends from successive sets of data. Once a Neural Network is optimally trained, it is able to make remarkably accurate prediction or perform the task it is designed to do. Types of Neural Networks There are various types of neural networks; the most popular type being the back propagation neural network (BPNN). BPNN is particularly adept at pattern matching and trend analysis. A conventional BPNN uses three layers of nodes, but it can use more middle layers. The first layer, the input nodes, receives the input data. The results of the first layer are passed to the next layer, which is also called the hidden layer. There can be more than one hidden layer, depending on the complexity of the problem at hand. This process of data propagation through the layers is repeated for each layer until an output is generated. The difference between the generated output and a training set output is calculated and then fed back to the neural network. Such feedback is used for adjusting the connection weights. This iterative process is meant to minimize the difference to within a predefined tolerance to reach the stage whereby the neural network has been optimally trained and ready for use. Since neural networks are particularly good for trend analysis, they can also be used for analyzing price action of the Forex market or other financial markets. Additionally, a neural network can also be used to discover patterns in the data for optimal combination of indicators that yield good entry and exit signals. 
An example of a multiple neural networks for analyzing data and producing market forecasts Neural network technology is being employed to generate remarkably accurate signals. Through the training of these neural networks and their ability to adapt to changing market conditions, the trading signals remain accurate over time. As new data become available, the system makes predictions and corrects itself according to how close its predictions were. In essence, the system is able to learn continuously in an adaptive manner. It is worth emphasizing that it is NOT systems with the highest accuracy that are desirable. For one can also curve-fit past data to yield superior back-test results. However, such system is bound to fail miserably when subject to real-time market dynamics and behaviors. It is the most adaptive systems that thrive and survive. | 
