> ML_ALGORITHM // NON-NEGATIVE-MATRIX-FACTORIZATION_v1.0
Non-Negative Matrix Factorization (NMF)
Matrix decomposition method that factorizes non-negative multivariate data into non-negative basis and coefficient matrices.
Matrix Decompositionclassical-unsupervisedhigh-intrinsicmedium (1k-100k)
Back to All AlgorithmsComputational Complexity
Training Complexity:O(iterations * n * p * k)
Inference Complexity:O(p * k)
Hardware Profile
CPU Friendly:Yes
Requires GPU:No
Memory Footprint:low
Interpretability & Data
Interpretability Tier:high-intrinsic
Training Data Needs:medium (1k-100k)
Interpretability Assessment
Components are non-negative and directly interpretable as additive building blocks of the data.
Suitable Tasks & Supported Modalities
Suitable Tasks:
feature extractiontopic modelingdimensionality reduction
Supported Modalities:
tabulartextimage
Implementing Libraries
Foundational Literature
Common Pitfalls & Warnings
- Attempting to fit data with negative values will throw runtime errors without prior non-negative translation