This research paper discusses Explainable Artificial Intelligence (XAI) techniques used in Machine Learning and their importance in understanding machine learning predictions. Machine learning models are increasingly used in different areas such as healthcare, finance, education, customer analytics and business decision-making. However, some advanced machine learning models can produce predictions without clearly showing how or why a particular decision was made. This creates a challenge because users may find it difficult to understand, verify or trust the output of the model.

This research focuses on exploring and understanding Explainable AI, different types of explanation techniques, their uses, challenges and possible solutions. It discusses commonly used techniques such as Local Interpretable Model-Agnostic Explanations (LIME), SHapley Additive exPlanations (SHAP), Partial Dependence Plots (PDP), permutation feature importance and surrogate models. XAI techniques can help users understand the contribution of different features to machine learning predictions. SHAP, for example, provides feature contribution explanations, while LIME explains individual predictions by approximating a complex model locally.

The study follows descriptive research based on secondary data collected from research papers, journals, academic publications and reliable online sources. The research aims to provide a better understanding of how XAI techniques can improve transparency, interpretability and understanding of machine learning models while also identifying their limitations and challenges.

Keywords: Explainable AI, XAI, Machine Learning, Interpretability, LIME, SHAP, Model Explainability, Artificial Intelligence