In addition to the price of the commodity, we should also try to obtain and maybe predict the volume traded.
One of the ideas of using neural networks to predict the price is the ability to capture the underlying trading activities of the market. Each neuron can represent a single or groups of similar traders, when a certain piece of the information is present, the neuron is activated and mimics the trading mechanism in the market.
If we can also predict the volume traded with high accuracy, we can interpret the neural network as a simplified and summarised version of the true market dynamics.
In addition to the price of the commodity, we should also try to obtain and maybe predict the volume traded.
One of the ideas of using neural networks to predict the price is the ability to capture the underlying trading activities of the market. Each neuron can represent a single or groups of similar traders, when a certain piece of the information is present, the neuron is activated and mimics the trading mechanism in the market.
If we can also predict the volume traded with high accuracy, we can interpret the neural network as a simplified and summarised version of the true market dynamics.