Machine Learning Approach to Forecast the Tensile Strength of Bamboo

Abstract

Bamboo holds significant global economic importance as a non-timber forest product. Its fibers can be utilized in concrete reinforcement, a concept known as bamboo fiber reinforced concrete. Nevertheless, further investigation is required for Bamboo Fiber Reinforced Concrete, with a specific focus on bamboo tensile strength. This study aimed to foretell the BTS through several machine learning techniques, including artificial neural network, extreme learning machine, and support vector regression. A total of 30 samples from the previous literature article were considered for predicting Bamboo Tensile Strength. The dataset was split into two parts, with 80% allocated for training data and 20% for testing data. The outcome from the data testing shows that the extreme learning machine predict very sensitive to random errors in the observed target. The results show that a positive correlation was found between key input parameters, such as the shorter dimension of bamboo, the longer dimension of bamboo, cross- sectional area, modulus of elasticity, and area, in relation to the output bamboo tensile strength.