From Galton to AI: A Data-Driven Approach to Height Prediction Using Growth Curve Comparison and Parental Data

Accurately predicting a child's height based on genetic and environmental factors is a longstanding challenge in medical and data science research. Traditional models, such as linear regression, SVM, and Random Forest, have been widely used. However, their accuracy varies due to the complexity of the environment, genetics which affects the height. This paper investigates the possibility of machine learning techniques, including deep learning techniques, accompanied by traditional models, to improve prediction accuracy for height. By combining more features and systematically partitioning datasets, the study compares various algorithms perform in predicting child height. These findings provide a comparative evaluation of these models, demonstrating the impact of feature selection and model architecture on prediction performance. The study aims to offer insights into selecting optimal models for practical applications in genetics, health analytics, and predictive modelling.