Hi Wafaa,
From my understanding of your question, you are looking to perform feature reduction, and although you have attempted using PCA, it has not yielded the desired results. Reducing dimensionality is often crucial for improving computational efficiency and potentially enhancing model performance.
Here are some alternative dimensionality reduction techniques you might consider:
- t-Distributed Stochastic Neighbor Embedding (t-SNE): A nonlinear technique that visualizes high-dimensional data by reducing it to two or three dimensions, preserving local similarities.
- Linear Discriminant Analysis (LDA): A linear technique that reduces dimensionality while preserving class separability by maximizing the distance between classes.
- Independent Component Analysis (ICA): A statistical method that separates a multivariate signal into additive, independent components, often used to identify underlying factors.
Additionally, you might explore using autoencoders. These are neural networks designed to learn efficient encodings of input data and can be utilized to compress features into a lower-dimensional space.
I hope these suggestions prove helpful for your task!
You can refer to this article for more detailed information:
