Tensorflow
- Google made an open source library for Machine Learning and Deep Learning.
- Data flow graph: Directed Graph using Node and Edge
- Node: Calculation for mathmatics, Data Input/output, Data Reading/Saving
- Edge: The relationship of nodes. Carry a tensors (dynamic sized multiple data array)
Terms
- Operation(op): Node in graph called op. op gets over than one tensor.
- Tensor
- Session: Capsulation of execution environment of the op
- Variables: Buffer for saving tensor in memory. Handle parameters when graph execution time
Referense
'배움의공간(學) > 데이터분석' 카테고리의 다른 글
YOLO (You Only Look One) (0) | 2024.03.07 |
---|---|
ADsP 준비 (0) | 2020.07.06 |
구글 머신러닝 집중과정 (0) | 2019.09.03 |
삼성SDS, "빅데이터 허브 되겠다" - Brightics Studio (0) | 2019.05.27 |
댓글