What to do after preparing a Model?

Asked

Viewed 67 times

1

I would like to understand how I can put into practice in some system, the model I create with Deep Learning or machine Learning. I see several tutorials teaching to create a very simple structure of deep learning, and in the end bringing results of approximately 90% acurracy. I understand that "it worked".

But now what? How to use this? How to put this model in my system?

machine-learning deep-learning

1 answer

1


There are some ways, I would say that the following two are the most used.

1) API:

There are several examples of API’s that provide model drainage (see: https://cloud.google.com/vision/). Create an endpoint that receives a new observation and returns the value predicted by the algorithm.

2) Strut in batch

The model is propped daily, for example, for all new observations.

Given these two ways there are some frameworks that help to do this without having to worry about scalability, etc. Are these:

In practice, it is easier to make an API in any language to do this than to use some framework..

  • Show. In the case of the Vision API, I can create my templates without necessarily being image related?

  • 1

    In this case, the Vision API is an example model that has been made available in API form. At first, any model can be made available in the form of API... You receive the information that is used for scoring and return the predicted result.

Browser other questions tagged machine-learning deep-learning

You are not signed in. Login or sign up in order to post.