PCA vs. t-SNE
...a popular interview question.
RouteLLM: Choose the Best LLM for Your Query!
Use a single interface to access all SOTA LLMs instead of paying separately for all of them:
OpenAI o1 for hard problems.
Claude Sonnet 3.5 for standard coding.
OpenAI GPT-4o for essays.
Gemini Flash 2.0 for longer contexts.
Mix and match to use the best LLM for your query on ChatLLM →
Thanks to Abacus AI for partnering today.
PCA vs. t-SNE
PCA vs t-SNE is a popular data science interview question so let’s understand how they differ today.
This table summarizes the differences between them:
Let’s discuss the differences below.
#1) Purpose
PCA is primarily a dimensionality reduction algorithm. It is NOT inherently designed to create 2D visualizations of high-dimensional dataset.
t-SNE, however, is a data visualization algorithm. We use it to project high-dimensional data to low dimensions (primarily 2D).
#2) Type of algorithm
PCA is a deterministic algorithm—running it twice on the same dataset will produce the same result.
t-SNE is a stochastic algorithm—running the algorithm can produce entirely different results. Can you explain why? We covered it pretty recently.
#3) Uniqueness of solution
As far as uniqueness and interpretation of results is concerned…
PCA always has a unique solution for the projection of data points.
t-SNE, as discussed above, can provide entirely different results, and its interpretation is subjective in nature.
#4) Projection type
PCA is a linear dimensionality reduction approach. It can only find a linear subspace to project the given dataset. KernelPCA addresses this:
t-SNE is a non-linear approach. It can handle non-linear datasets.
#5) Projection approach
PCA retains the global variance of the data. Thus, local relationships (such as clusters) are often lost after projection, as shown below:
t-SNE preserves local relationships. Thus, data points in a cluster in the high-dimensional space are much more likely to lie together in the low-dimensional space.
Note: In t-SNE, we do not explicitly specify global structure preservation. But it typically does create well-separated clusters.
That said, the distance between two clusters in low-dimensional space is NEVER indicative of cluster separation in high-dimensional space, as depicted below:
Those are the key differences between PCA and t-SNE.
And before we end...
Don't forget to check out ​ChatLLM​ to access all SOTA LLMs in a single interface instead of paying separately for each of them:
👉 Over to you: What other differences between t-SNE and PCA did I miss?
Thanks for reading!