Machine learning has come a long way since early 1940's. Everyday I read so many papers about new improved models that has better accuracy/lower error rate compared to the one's that already exist. One of the major consumers of these models are industries. My question is going to have be of two parts
1. Have you programmed your ML model in one of the most popular languages such python & R and shared it in Git?
2. When you proposed a model or a new techniques or a workflow, have you ever considered a feature on implementing this in production environment? If so, how did you do it?
Python is the only language I use for ML/AI works.
Regarding the deployment, I think the question should be more related with the 3V of the data. I have seen applications deployed on cloud with huge clusters, pipelines and high amount computing power to process data coming like a rain (talking about streaming for every milliseconds). Otherwise, I think for R&D and demonstration purposes, an interactive notebook hosted on cloud might be more than enough.
Hi,
If you are searching for ML examples programmed in R of Python AND examples of practical implementations have a look at Kaggle:
https://www.kaggle.com/
They have competitions where people compete to provide the best implementation of ML for practical applications (sometimes with the possibility of winning a lot of money) and they also have a "Kernel" section where people share their R/Python code. Not sure if any of these competitions/kernels are examples of how to implement ML in a production environment, but it might be an interesting starting point to search for such examples. And inspiration.
Best,
Urs
I am not an expert..but just started using Apache Spark, to create a production model for a simple Polynomial regression (which I had experimented with in scikit-learn) to predict output from streaming data coming from an IOT device (raspberry pi for prototyping).
I'm using python for ML model development and for the deployment the Transfer Learning can be seen as a ‘design methodology’ within Machine Learning like for example, active learning.
- Badges
- Science topic
- Similar topics
- Papers