Explanation of RNN algorithm.

Recurrent neural networks (RNNs) are a type of artificial neural network that is designed to process sequences of data, such as text, speech, and time-series data. They are designed to have a "memory" in order to learn from previously seen sequences of data. Unlike other types of neural networks, RNNs contain neurons that have self-loops or “recurrent” connections. These loops enable the networks to remember the sequence in which it sees data points, allowing the neural network to capture information in a more meaningful way. 

RNNs take an input, like text or a time-series signal, and make predictions based on what it has seen before. It begins with a single input data point and passes it through the network of neurons in the model. At each step, the model takes the previous input, combines it with the current input, and processes the two to produce a new output. This output is then fed into the next layer in the network, which combines the current input with the previous output, producing a new output. This process is repeated for every data point in the sequence, resulting in a more complete understanding of the sequence of data.