This thesis studies the theoretical and experimental determination of the fractal dimension of different sets. It contains both pure and applied topics. After establishing the ground rules, we look at ways to calculate the fractal dimension of fractal interpolation functions and we end the study with an experiment consisting in the estimation of the fractal dimension of two graphs representing the activities of two companies at the Toronto Stock Exchange. In fact, we will see that there is a certain cohesion between the fractal dimension of a cloud of points and the absolute value of the slope of the regression line approximating the same cloud of points.