We only had forward connections in neural network models up to this point. In this chapter, we will add feedback connection between neurons to form recurrence. This will help us in forming neural network structures with memory capability. Hence, they will be more suitable for sequential data processing. To explain these concepts better, we will first associate recurrence with memory in this chapter. Afterward, we will introduce three popular neural network models based on recurrence as recurrent neural networks (RNN), gated recurrent unit (GRU), and long short-term memory (LSTM). While introducing each model, we will first provide general information about it. Then, we will focus on its formation and training in Keras. Then, we will provide the usage examples of recurrence models on PC. Afterward, we will consider implementing recurrence models on the STM32 microcontroller. Finally, we will provide examples on the usage of recurrence based models to solve real-life problems.