I found a post in stakeoverflow here What is the algorithm behind Rasa NLU? - Stack Overflow mentioning that Rasa NLU doesn’t use any algorithms, it uses specific piplines which you specify in the config.yml. So a component in the pipeline is an algorithm? If not, what algorithm does Rasa NLU or Rasa uses?
@AhChing11 Heya! Nice and simple questions Let me explain you.
What is Pipeline?
A Pipeline is a set of data processing units arranged in series such that the output of one element is the input of the subsequent element. Pipelining is a technique in which multiple instructions are overlapped during execution. It is used primarily to create and organize instructions in a processor so that the processes run in cocurrent fashion.
What is Algorithm?
An algorithm is a set of instructions for solving a problem or accomplishing a task .
Rasa uses:
- Pipeline: Tuning Your NLU Model
- Alogrithm Whiteboard: NLP for developers YouTube series! - #3 by Emma
- Alogrithm Whiteboard: Introducing The Algorithm Whiteboard! - #4 by Emma
- Alogrithm Whiteboard Youtube: Introducing The Algorithm Whiteboard - YouTube. [all are explain in detail, what all algorithms are used by rasa ]
Extra:
With Algorithmia, pipelining machine learning is simple:
- Algorithms are packaged as microservices with API endpoints: calling any algorithm or function is as easy as
algorithm.pipe(input)
- Pipelines can be input agnostic, since multiple languages and frameworks can be pipelined together
- You can set permissions for models and choose to allow a model to call other algorithms
I hope this will solve your question use the concept of pipelines vs algorithm. Good Luck and do see all the ref link.
That’s a little much to process at once XD. I’ll slowly read those ref links, thanks for now :D.
@nik202 Does the custom component in the pipeline can classify as a machine model as well e.g adding sentiment analysis model?