Abstract: Crime analysis methods are used to analyze the crime data for finding different data trends and patterns in the crime. Crime can be predicted using different models. Crime analysis can find the trend of crime type in different locations. As the crime increases, the data to be processed also increases. The ability to predict the crimes based on the location, sample, and time can serve as a valuable source of information. To predict crime correctly with higher overall performance is a hard task due to the increasing numbers of crime. Therefore, crime prediction technique is critical to discover the future crime and thus to decrease the numbers of crime. In this survey, we are discussing about the overall performance of prediction models such as ARIMA model, Recurrent Neural Network, Long Short Term Memory which are used for crime analysis and prediction.