Article

Exploring Artificial Intelligence for Learning Enhancement with Predictive and Explainable Modelling

---

Abstract

This study explores the application of Artificial Intelligence (AI) in learning through the use of a Linear Regression model to analyse and predict relationships among financial dataset variables. The dataset consists of several numerical features, including open, high, low, close, and trading volumes for XRP and USDT. Descriptive statistics, correlation analysis, and data distribution visualizations were conducted to provide an understanding of the dataset before implementing the predictive model. The correlation results showed very strong linear relationships among the price variables (r ≈ 0.99) and a strong positive correlation between trading volumes (r = 0.82), indicating highly synchronized market movements. The AI model achieved exceptional predictive performance with an R² value of 0.993, a Mean Absolute Error (MAE) of 0.0099, and a Root Mean Squared Error (RMSE) of 0.0286, demonstrating that even a simple algorithm can accurately capture complex numerical patterns. The feature importance analysis revealed that the high and low variables were the most influential predictors of the closing price, providing a clear example of model interpretability and eXplainable AI (XAI). From an educational perspective, this research illustrates how AI can serve as a practical and interactive learning tool that helps students understand core data science concepts, including data preprocessing, correlation, model evaluation, and interpretability. The study concludes that integrating AI into learning environments enhances students’ analytical thinking, promotes data literacy, and encourages a deeper understanding of how AI systems learn and make predictions.

Keywords: Artificial Intelligence, Machine Learning, Predictive Modelling, XAI, AI in Education

How to Cite:

Thepnuan, N., (2026) “Exploring Artificial Intelligence for Learning Enhancement with Predictive and Explainable Modelling”, Artificial Intelligence in Learning 2(1). doi: https://doi.org//AIL.121