How to use JPMML-Android to implement a PMML machine learning model?

Question

I've created a PMML Random Forest model using scikit-learn in Python, and my goal is to implement that model in an Android app to do live predictions. From a previous question, the best way to go about this seemed to be using JPMML-Android ( https://github.com/jpmml/jpmml-android ). However, there's no tutorials, and given how new I am to this, I'm not sure at all what's going on in that 'Usage' section. Any help would be greatly appreciated on how I can use this to implement my model!

Thanks!

Answer 1

1. put your .pmml file in the directory:

jpmml-android-master\\pmml-android-example\\src\\main\\pmml\\

2. install the mvn plugin: eg in windows cmd:

mvn clean install

3. find your .ser file in the directory:

jpmml-android-master\\pmml-android-example\\target\\generated-sources\\combined-assets\\

The 2nd step compile the code and convert the .pmml file to .ser file.

Then you can follow the example code in

jpmml-android-master\\pmml-android

to build the android project in Android Studio to use the .ser model.