What technique would you use to fill in missing data values between existing data points?

Interpolation is a technique used to estimate missing data values by leveraging the known values that surround the missing point. It assumes a certain continuity in the data and enables filling gaps in a dataset based on the trend or pattern observed from existing values. This approach is particularly useful when dealing with time-series data or ordered data where the relationship between values is sequential and can be approximated.

In contrast, imputation generally refers to a broader strategy that encompasses techniques for filling in missing values based on various algorithms, which can include mean, median, or mode of the existing values, but not specifically by using interpolation methods. Regression focuses on modeling the relationships between variables, while normalization is concerned with scaling data to a specific range, such as converting to a 0-1 range. These techniques do not specifically address the task of filling in missing values based on adjacent data points, which is the core purpose of interpolation.