What is the process called that simplifies a dataset by removing redundant or irrelevant features?

Dimensionality reduction is the process focused on simplifying a dataset by eliminating redundant or irrelevant features. This practice is important in data science and machine learning because it helps improve model performance by reducing overfitting, decreasing computational costs, and enhancing the interpretability of the model.

Through techniques such as Principal Component Analysis (PCA) or t-Distributed Stochastic Neighbor Embedding (t-SNE), dimensionality reduction enables the transformation of a large set of variables into a smaller set while retaining as much information as possible. By doing this, it not only streamlines the dataset but also focuses the analysis on the essential features that contribute the most to the variability in the data.

In contrast, data cleaning refers to the process of identifying and correcting errors or inconsistencies within the data, which is more about quality than feature selection. Data wrangling involves transforming and mapping raw data into a more usable format but doesn't necessarily focus on reducing feature space. Feature engineering involves creating new features from existing ones to enhance the model's predictive power but does not eliminate features. Thus, dimensionality reduction specifically targets the simplification of a dataset by focusing on the removal of unnecessary features.