Support Questions
Find answers, ask questions, and share your expertise
Announcements
Alert: Welcome to the Unified Cloudera Community. Former HCC members be sure to read and learn how to activate your account here.

Deployment of machine learning models

Deployment of machine learning models

New Contributor

Hello,

 

Could some1 suggest good resources for deployment machine learning model in Hortonworks (Cloudera) on premise Data Lake ? 

 

I have working model for upsale prediction, and works on my local computer when i run it manually it takes data from WareHouse. I want deploy solution to Data Lake were exist same information (in terms of tables and data).

 

What I have tried, but did not work your is to upload to python file to Data Lake server , try make connection to Hive, extract data and run model. Could not connect to Hive (but that not what I looking for). 

 

I rather looking for help in term best practice to deploy models in Data Lake. any help will be appreciated. 

3 REPLIES 3
Highlighted

Re: Deployment of machine learning models

Super Collaborator

@Stass I don't think you've provided enough information so that members of the community can provide a useful response to your question. What framework are you using for building the machine learning models on your local computer? What environment are you using to support the models for deployment?

 

 

 

Bill Brooks, Community Manager
Was your question answered? Make sure to mark the answer as the accepted solution.
If you find a reply useful, say thanks by clicking on the thumbs up button.
Highlighted

Machine Learning model deployment

New Contributor

Hello,

 

I would like aquire information about how can I can deploy Machine Learning project on Hortonworks in perment Data Lake.

 

Data is stored in Hive.

 

I am working with python on my local computer and have developed model. Now I looking for solution to deploy it.

 

I have tried to write beeline command that is executed in python to connect to Hive database and extract data ( but did not work out) connection issue and firewall. Looking for other suggestions, maybe Zepellin + Ambari combination or maybe any ideas about Dockers.. Any help would be appreciated. 

Highlighted

Re: Machine Learning model deployment

Super Collaborator

@Stass What framework are you using for building the machine learning models on your local computer?

Assuming that you are looking for recommendations on an environment to use to support the models for deployment, you should know that Hortonworks, before it merged with Cloudera, didn't have a ML product. The only real solution they had was via their partnership with IBM, which offered it's Watson Studio line of products. Using other projects, such as TensorFlow, you could deploy ML models if you were using such frameworks.

Cloudera has a ML environment called Cloudera Data Science Workbench which now works with Hortonworks' Data Platform (HDP). If you're doing any sort of  collaboration with others at your company, you should definitely look into that product. 

 

 

Bill Brooks, Community Manager
Was your question answered? Make sure to mark the answer as the accepted solution.
If you find a reply useful, say thanks by clicking on the thumbs up button.
Don't have an account?
Coming from Hortonworks? Activate your account here