Today 90% of the Machine Learning Models does not succeed because of the reason that whenever you change something into the code, after training it, to improve it's performance and accuracy, and if our model is build using BigData then it will take a lot of time, computing power and resources to train our model again and if after then the accuracy does not matches our requirements then you again need to tweak the model and wait for it to again get trained.
So to resolve this problem we need to come up with the idea of integrating Machine Learning with DevOps i.e. "MLOps" so that we can automate some of the tasks of Data Scientist like tweaking of model
- Python Language
- Should know about the Machine Learning concepts for building the model.
- To automate the model you should know about the concepts of DevOps like Jenkins, GIT , Github, etc.
- Docker