```
tags:
- AI/Tasks/DimensionalityReduction
aliases:
- Dimensionality Reduction Techniques
- Feature Reduction
```

Dimensionality Reduction Technique(Feature Reduction) is the process of reducing the number of features in a resource heavy computation without losing important information. it reduces model complexity and Overfitting at the cost of accuracy.

Dimensionality Reduction is part of two processes:

- Feature Selection: select a subset of features using Feature Selection Techniques. It perform dimensionality reduction by disregarding less valuable features.
- Feature Extraction: extracting or deriving information from the original features. It perform dimensionality reduction by disregarding many features in favor of fewer, more valuable features.

Info

Dimensionality Reduction is a response to the problems caused by the Curse of Dimensionality, such as increased complexity of models, increased errors, and need for more data.

Algorithms:

- Principal Components Analysis (PCA)
- Singular Value Decomposition (SVD)
- Linear Discriminant Analysis (LDA)
- Quadratic Discriminant Analysis (QDA)
- Autoencoders (AE)
- Independent Component Analysis (ICA)
- Generalized Discriminant Analysis (GDA)
- Non-negative matrix factorization
- t-Distributed Stochastic Neighbor Embedding (t-SNE)
- Fourier and Wavelet Transforms
- Multi-Dimensional Scaling (MDS)
- Canonical Correlation Analysis (CCA)
- Fisher’s Linear Discriminant (FLD)
- Partial Least Squares Discriminant Analysis (PLSDA)
- Mixture Discriminant Analysis (MDA)
- Regularized Discriminant Analysis (RDA)
- Flexible Discriminant Analysis (FDA)
- Projection Pursuit
- Sammon Mapping

Interactive Graph

Table Of Contents