In this study, while reviewing statistical methods for data collection and statistical surveys, the importance and necessity of using data science methods to call official and registered data/online data is discussed. Then, with the help of Application Programming Interface (API), we will examine how to call corona virus (COVID-19) data as well as crypto-currency data from different databases. Then using various time series statistical methods including ARIMA, NNETAR, ARFIMA, Holt-Winter, Hybrid, BSTS, TBATS, Prophet and neural network models such as MLP and ELM try to fit online models on the data and after training and testing the modeling results and comparing them with each other by using goodness of fit measures as well as residual measures, the best model is selected. Then, based on the selected model, forecasts are made for the next time periods, and decisions are made about the next step, including buying or selling or keeping the crypto-currency in the digital wallet and the action is taken online. In fact, the entire above are prepared as a robot to perform the necessary and appropriate actions at the same time. It should be noted that some of the above cases have been done by R software and some with the help of Python programming language. For more appropriate and easier use of the above, the relevant codes are converted into appropriate applications using the well-known Shiny package in R software.

Keywords