Sparse Matrix
In plain terms
A matrix that's mostly zeros, stored compactly.
Going deeper
A sparse matrix has so few non-zero entries that storing only those (with their positions) saves huge memory and speeds computation. Special formats avoid wasting space on zeros.
Under the hood
Formats like CSR/COO store non-zeros and indices; essential in scientific computing and ML where data is high-dimensional but sparse.
Examples
Recommendation systems, graph adjacency, NLP feature vectors, finite-element analysis.
Think of it like...
Like noting only the few occupied seats in a giant empty stadium instead of every seat.