This research aims to study: the use of matrices in predicting the future value of a simple linear regression model, by identifying the processes related to matrices and their concept, as well as touching on the concept of linear regression as a "linear equation" and methods of predicting it and its estimating properties, and the researcher reached a set of results: Method of predicting the use of matrices is one of the methods of predicting the future value, and there are many applications related to matrices in other mathematical sciences, and prediction contributes in many scientific fields, including administrative, economic and mathematical, and access is a simple and easy way to predict the future value using matrices, and that methods and methods of prediction generally assume That the underlying factors present in the past will continue in the future, and this represents the tendency of phenomena to be repeated in the future, and that predictions are rarely complete; Actual results are usually different from estimated or predicted values, and the inability to accurately predict is due to the multiplicity and multiplicity of the variables affecting or the influence of random factors; Therefore, the limits of variation and the extent of deviation are set to take these factors into consideration, and the accuracy of the prediction decreases the longer the time horizon for forecasting, and generally short-term forecasts are more accurate than long-term predictions: because the first is less likely to be uncertain of the second, the historical data that usually forms time series What takes a certain form is called a change pattern, and knowing the latter helps to achieve more accurate predictions. As for historical data characterized by an unstable and stable pattern of change, the pattern is hidden and unclear; it does not help achieve accurate forecasts and its prediction errors are significant.