Skip to content

Label Propagation Algorithm by Rust. Label propagation (LP) is graph-based semi-supervised learning (SSL). LGC and CAMLP have been implemented.

License

Notifications You must be signed in to change notification settings

vaaaaanquish/label-propagation-rs

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

label-propagation-rs

Label Propagation Algorithm by Rust.

Label propagation (LP) is graph-based semi-supervised learning (SSL).

A simple LGC and a more advanced CAMLP have been implemented.

Usage

You can find the examples in the examples directory.

The label is a continuous value of [0, class_n], and the result of predict_proba is the value of the label.

use std::result::Result;
use std::error::Error;

extern crate label_propagation;
extern crate ndarray;
extern crate ndarray_stats;

use ndarray::prelude::*;
use label_propagation::{CAMLP, LGC};
use ndarray::Array;

pub fn main() -> Result<(), Box<dyn Error>> {
    // make graph matrix ndarray
    let graph = Array::from_shape_vec(
        (3, 3),
        vec![0.0, 0.3, 0.0, 0.3, 0.0, 0.0, 0.0, 0.0, 0.0]).unwrap();

    // node index for label
    let x = array![0, 1];
    // label
    let y = array![0, 1];

    // make model
    let mut model = CAMLP::new(graph).iter(30).beta(0.1);
    // let mut model = LGC::new(graph).iter(30).alpha(0.99);

    model.fit(&x, &y)?;

    let target = array![0, 1];
    let result = model.predict_proba(&target);
    println!("{:?}", result);

    Ok(())
}

develop

docker build -t graph .
docker run -it -v $PWD:/app graph bash

Thanks