The human, ever since his emergence on the Earth, has always wanted to know what the future would bring, what events could happen. People wanted to know this not out of idle curiosity, but to be better prepared for these events. That’s the way forecasting appeared. Currently, there are different kinds of forecasts. Forecasts can be divided into short-term, middle-term and long-term. They can also be individual, local, regional, etc. But whatever be the forecast, it is based on a forecasting model, i.e. the tool which is used for forecasting. The present paper is devoted to the analysis of the main models used for time series forecasting. The paper deals with the following types of forecasting models: regression and autoregression models, exponential smoothing models, neural network models, Markov chain models, models based on classification and regression trees, models based on the genetic algorithm, support vector and transfer function models, fuzzy logic models, singular spectrum analysis models, local approximation models, models based on fractal time series, models based on wavelet transformation, models based on Fourier transformation. Along with studying the structure or algorithm of each model, the paper also attempts to identify their strengths and weaknesses. © Research India Publications.