The Verification of Propotionality Between Time Period and Length of a Simple Pendulum Experiment Using Deep Neural Network

Abstract

This research shows a pedagogic experimental and theoretical study of the motion of a simple pendulum, which considers the propotionality to the variables length (L) and period-time (T) of a simple harmonic motion, is presented. The study has used RELU (RECTIFIED LINEAR UNIT) activation function in deep neural network which is a branch of artificial neural network to examine the correlation between the dependent and independent variables in a simple pendulum experiment, the variables and their values was first generated from an online physics laboratory, the values and their corresponding variables were later separated into two CSV files, after which they were analyzed with the use of linear regression model in PYTHON programming language. It also applies to Physics Direct Method to represent these equations, in addition to the numerical solutions discusses. This research investigates the relationship between Length and Period using neural network models to find out a unique numerical solution by using simulation to see their behavior which shows in last part of this article. The results obtained shows that the linear approach to modeling the relationship between a scalar response and one or more variables with the RELU activation function proves their proportionality, this would be a good reference against which other results obtained from other simple harmonic motion experiments can be compared.