Hacker News Reader

Comment by Irishsteve | Hacker News Reader

Top point talks about targeting data scientists. It would be interesting seeing what that means in their mind.

Based on the current state of affairs oracle and the jvm seem very far away from mainstream ML

There's a lot of enterprise data science being done with Java/Scala and the Hadoop/Spark ecosystem.

MXNet also makes deep learning a first-class citizen in Java, but yes the research community is firmly entrenched in Python atm.

I see JDK languages as being in a decent spot for deploying ML/DL, and maybe an "emerging" language for training DL models.

Folks like Jeremy Howard have increasingly been expressing growing pains with Python and TensorFlow has been looking for something with a better type system like Swift.

If I had to take a guess, I'd say that Python is likely going to be sharing the ecosystem with a language with a better type system and performance like Julia or something on the JDK.

The Tensorflow Java project is alive and well. The next version of the API based on TF 2 is coming out soon. (Full disclosure, I'm a member of the SIG that's building it). I'm a firm believer in the utility of type systems for building machine learning projects.

https://github.com/tensorflow/java

Can you train models with TF Java? I was under the impression it was for inference only

You can with some hoop jumping in the current release (e.g. using 1.14 you need to define the model and optimizer in Python - https://github.com/oracle/tribuo/blob/main/Interop/Tensorflo...). With the upcoming alpha release from the TF Java group you can train purely from Java without invoking any Python - https://github.com/tensorflow/java-models/blob/master/tensor...

I too would like to know what's going on in their mind. Given they're apparently under the impression there that the most notable contender in that space is Go. Scala isn't a bad call though, and is JVM. Java itself seems like a non-starter for doing actual data science due to its verbosity and lack of scripting.

Instead, Julia should really be first to mind when considering that question. Although I might expect the author to respond with something about the mysteries of 'production'.

A lot of that territory is well worn by various Java libraries. Not a lot of money to be made so Oracle is probably better off not bothering with anything ML or data science related, nothing here to see that would interest them in any way, nope.

Actually my group in Oracle Labs launched a Java ML library last week - https://tribuo.org.

Nobody in their right mind would choose Java for anything ML-related.

Unless there were doing fairly simple ML that needed to be tightly integrated with a pre-existing Java codebase.

Can you elaborate on why you feel that way?

I will risk that I get downvoted to hell. The data scientists in my company and pure geniuses. But. They can barely program. I wouldn’t trust them with Java, I’m sure they wouldn’t trust themselves either. Python is slow but at least it’s easy and fun. I don’t see anything else getting traction for long years.

That’s too broad a statement. Can you substantiate why you think so?

The book below provides a nice overview of using Java for DS and ML

https://learning.oreilly.com/library/view/data-science-with/...

Seems a bit of an overstatement - there are huge production systems doing massive ML using Spark, Hadoop, Kafka etc. In some areas these are the defacto solutions. They can't all be "not in their right mind".

Reply on news.ycombinator.com