What is the difference between artificial neural network and hidden Markov model methods?

Answer 1

An artificial neural network is a structure which will attempt to find a relationship i.e. a function between the inputs, and the provided output(s), in order that when the net be provided with unseen inputs, and according with the recorded internal data (named "weights"), will try to find a correct answer for the new inputs. Hidden Markov models, are used for find the states for which a given stochastic process went through. The main difference could be this: In order to use a markov chain, the process must depend only in it´s last state. For use a neural network, you need a lot of past data. After training process, neural networks are capable of predicting next states of the system based only on the last state. In addition, given the ability to measure the prediction error (for example, after actual event, signal or state has happend and was compared to prediction), the neural network is capable of adapting itself and capture online changes in the undergoing process to improve the model of prediction and decrease the estimation error for the next states. Theoretically such approach can eliminate the need in initial training, as the network started from some random model will eventually adapt itself to the actual process it tries to estimate given this feedback error loop and will start to make correct estimations / predictions after a certain amount of steps. In such setup one can assume that neural network can be used when no past data is available at all. In this case neural network build the model of the ongoing process "from scratch" based on the observations in the "online" mode.

Answer 2

Artificial neural network is a machine learning model that processes data in a non-linear way, while hidden Markov model is a statistical model that deals with sequential data and uses probabilities to model transitions between states. In neural networks, the focus is on learning the underlying patterns within the data, while in hidden Markov models, the emphasis is on estimating the probabilities of transitioning between different states.