Definition
Data Fitting is the process of finding a mathematical function (a model) that best represents the trend in a given set of data points.
Why It Matters
Data fitting is the essential tool of machine learning and finance, enabling us to generalize from the past to bet on the future.
Core Concepts
- Linear Regression: The most common form of data fitting, which uses the “least squares” method to find a line that minimizes the sum of the squares of the vertical distances from the data points to the line.
- How to read: “The equation y equals m x plus b.”
- Meaning: Best-fit straight line through scattered data—m is slope, b is y-intercept; chosen to minimize squared vertical errors.
- Curve Fitting: Extending the concept to non-linear models, such as polynomial, exponential, or power functions, to better fit data that does not follow a straight line.