Ever wondered how scientists manage massive data without endless storage?
Tensor network
Ever wondered how scientists manage massive data without endless storage?
Imagine trying to store a huge number of photos on your phone. As you take more, it gets harder to fit them all in without deleting some.
Think of your phone's storage like a big closet. Instead of trying to fit all your clothes (photos) in one room (storage), you organize them into smaller boxes (tensors) that you can stack neatly (contract). This way, you can fit more clothes without needing a bigger closet.
Example
If you have 1,024 photos and organize them into 32 boxes, instead of needing a huge closet, you just need 32 smaller ones.
Remember this
Tensor decomposition helps compress high-dimensional data into smaller, manageable parts.
Text adapted from Wikipedia, licensed under CC BY-SA 4.0.
Related concepts
tensor cores are
Why can computers crunch numbers faster than humans?
Singular value decomposition
Ever wondered how to compress images without losing quality?
Matrix (mathematics)
Ever wondered how computers can predict your favorite songs?
Vector quantization
How can you store a huge library of books in a tiny closet?
quantization to INT8 doubles throughput
Quantization to INT8 doubles throughput because tensor cores process INT8 2x faster
autoencoders learn the data manifold
Autoencoders compress data manifold by forcing information through a bottleneck layer, learning efficient representations
Swipe through 100 ML concepts daily
Open Pocket Polymath