ANNs essentially associate input patterns with output patterns. The inputs could be the raw stock market data, since this is the material that technical analysts use to predict movements in the market. The outputs could be any one of several things. For instance, given inputs representing the share prices on day 1, day 2 and day 3, the output might be a prediction of the share prices on day 4 (or possibly even on days 4 and 5). Alternatively, the outputs might be simple “buy” and “sell” signals for shares in particular companies.
The advantage of ANNs is that they are flexible. They will, if correctly trained, learn to classify any pattern in the training library correctly: if they are given a pattern on which they have been previously trained, they will produce the correct output. However, they will also produce the correct output if they are given a pattern similar to one they have seen during training. They will therefore classify patterns that they have never seen before, based on the closest matching training pattern.
Whereas conventional computer programs are written specifically and involve specific rules given by the human programmer, ANNs are trained by example. They learn to associate patterns on their inputs with corresponding patterns on their outputs. Training the neural net involves giving it a sample from the library of known input patterns and the corresponding desired output. Then the neural network is told to adapt its connections. This is repeated with other input-output pattern pairs. Typically, the set of input-output pairs used for training the net has to be presented repeatedly many times, and so neural networks can often take several hours to train. Once the neural network has been trained, you can present an input pattern to it and it will produce the corresponding output pattern.
How do neural networks work?
Whereas conventional computer programs are written specifically and involve specific rules given by the human programmer, ANNs are trained by example. They learn to associate patterns on their inputs with corresponding patterns on their outputs. Training the neural net involves giving it a sample from the library of known input patterns and the corresponding desired output. Then the neural network is told to adapt its connections. This is repeated with other input-output pattern pairs. Typically, the set of input-output pairs used for training the net has to be presented repeatedly many times, and so neural networks can often take several hours to train. Once the neural network has been trained, you can present an input pattern to it and it will produce the corresponding output pattern.
Neural Network Architectures
An ANN consists of a number of neurons simulated in a computer with connections between them. Each of these connections has a certain strength, or “weight”, indicating how strong the connection is. A connection with a large weight is efficient at passing on a signal from one neuron to another. The information and expertise that ANNs possess is coded in these strengths between neurons.
Simulated neurons can be connected in various patterns, termed “architectures”. Although complex architectures exist (for instance in Cell Assemblies, which model closely structures in the human brain), the simplest architecture, the Multi-Layer Perceptron, is probably the most common. An MLP consists of neurons in distinct layers, each of which is connected to the next layer along. Signals can pass only one way through the MLP, from the inputs of the first layer to the outputs of the last one.
What are Neural Networks?
Artificial Neural Networks (ANN) are simulations on a computer of groups of brain cells (neurons) configured in such a way that they perform a useful task. They are inspired by real neurons, which are simple switching elements – they gather electrochemical energy from their inputs, then either pass it on or block it. This means that individual neurons have little more intelligence than a light switch, and yet, but putting countless trillions of these cells together and setting up the correct connections between them, we get human intelligence and even (in a process that scientists haven’t even begun to understand) consciousness!
Technical Analysis does have its critics – those economists who say it is nothing more than a pseudoscience, one that has no basis in fact. Investors such as Warren Buffett and Peter Lynch say that there is precious little evidence that Technical Analysis works. However, various academic studies have shown that technical analysis, particularly when carried out by neural networks, can produce statistically significant returns, and many investors have made fortunes through the application of its principles. It is therefore unfair to dismiss it as mere nonsense, and one ignores it at one’s peril!
One of the tenets of Technical Analysis is that prices follow trends, either upwards, downwards or sideways (when the price is generally holding steady – a “flat” price). If investors spot an upward trend in a share, they may well try to jump on the band-wagon and buy that share, hoping to benefit from its rise, and thereby driving its price up further. Similarly, a downward trend may cause investors to dump shares on the market, causing a further fall in the price. This can cause the price to “zig-zag” – rise and fall several times in line with investor confidence.
Technical analysis is often compared to Fundamental Analysis, which also tries to predict share price but on the basis of company assets, the general health of the company and other factors such as the amount of competition in the market that a company faces. Technical analysis, on the other hand, is concerned only with the behaviour of the market: The market’s opinion of a particular company is more important than that company’s underlying worth. The theory is that all the information about the company that you need to predict its share price is already encoded in that price in some way.
Technical analysis is the science (some might say art) of predicting the prices of share commodities based on their current price and past performance. Share prices wander up and down, changing many times every day. If you plotted the price of a share during the course of a particular time period (day, month or year), you would get a line that moves up and down sharply, like some demented saw blade. You might, however, be able to spot a general pattern underneath all that variation. The random variation is termed “noise” and the purpose of technical analysis is to remove that noise as far as possible.