Design of the Approximation Function of a Pedometer Based on Artificial Neural Network for the Healthy Life Style Promotion in Diabetic Patients

The present study describes the design of an Artificial Neural Network to synthesize the Approximation Function of a Pedometer for the Healthy Life Style Promotion. Experimentally, the approximation function is synthesized using three basic digital pedometers of low cost, these pedometers were calibrated with an advanced pedometer that calculates calories consumed and computes distance travelled with personal stride input. The synthesized approximation function by means of the designed neural network will allow to reply the calibration experiment for multiple patients with Diabetes Mellitus in Healthy Life Style promotion programs. Artificial Neural Networks have been developed for a wide variety of computational problems in cognition, pattern recognition, and decision making. The Healthy Life Style refer to adequate nutrient ingest, physical activity, time to rest, stress control, and a high self-esteem. The pedometer is a technological device that helps to control the physical activity in the diabetic patient. A brief description of the Artificial Neural Network designed to synthesize the Approximation Function, the obtained Artificial Neural Network structure and results in the Approximation Function synthesis for three patients are presented. The advantages and disadvantages of the method are discussed and our conclusions are presented.