Somekey terms and their explanations Machine Learning (ML)—A subset of AI involving algorithms that learn patterns from data and improve performance on a task without being explicitly programmed for each scenario. ML methods are commonly used in risk prediction, image analysis, and clinical outcome modelling. Features: The variables that contribute towards the model’s output. Feature space represents a multi-dimensional space where each dimension corresponds to a specific feature of the given data. Every patient represents a unique data point in feature space. The dimensionality of the feature space corresponds to the number of features in the model. Example: Imagine plotting patients on a graph where one axis is age and the other is blood pressure; age and blood pressure are features, the 2D graph is the feature space, each patient is a data point on it, and adding more features like heart rate or cholesterol would add more axes (dimensions) to this space. Distance function: In the feature space, the notion of distance allows us to quantify how similar or different our features are collectively in a dataset. The distance between the points helps us identify clusters of similar or different classes. Decision boundary is typically an equation (a line, curve, or surface) that uses several features together (e.g., age + blood sugar + BMI), and the model decides based on which side of that surface a data point lies. For complex ML models (e.g., decision trees, random forests, deep nets), the boundary can be an irregular, wiggly surface, but conceptually it is still “the place where the prediction flips from one class to another”. The algorithm learns to draw this boundary from the data and is not random.