Brain.js: Neural network output values over 1?

[vorticalbox]

What is wrong?

after training on iris data https://archive.ics.uci.edu/ml/datasets/Iris/ output of one iris is massively higher than the others

How do we replicate the issue?

download the csv of data

const brain = require('brain.js');
const net = new brain.NeuralNetwork();
const fs = require('fs');

const data = fs.readFileSync('./iris.csv').toString();
const trainingData = data.split('\r\n').map((d) => {
    const [slength, swidth, plength, pwidth, flower] = d.split(',');
    return {
        input: {
            slength: slength / 10,
            swidth: swidth / 10,
            plength: plength / 10,
            pwidth: pwidth / 10,
        },
        output: {
            [flower]: 1,
        },
    };
});
const trainResults = net.train(trainingData);
console.log(trainResults);
const result = net.run({ 
    slength: 0.51,
    swidth: 0.35,
    plength: 0.13999999999999999,
    pwidth: 0.1 });

console.log(result);

How important is this (1-5)?

3

Expected behavior

{ 'Iris-setosa': 0.9973204731941223,
  'Iris-versicolor': 0.003389077726751566,
  'Iris-virginica': 2.462011443240403e-13 }

should be more like

{ 'Iris-setosa': 0.9973204731941223,
  'Iris-versicolor': 0.003389077726751566,
  'Iris-virginica': 0.003 }

Other Comments

training results

{ error: 0.008838672393814171, iterations: 20000 }

[mubaidr]

I think you missed e-13 in the result?

'Iris-virginica': 2.462011443240403e-13

[vorticalbox]

The input I used in the .run is Iris-setosa and that came out as 0.99 almost perfect.

is there where the float is long i'm getting the notation rather than the number? if so is there anyway to get the network to just round so I don't get values like this?

[mubaidr]

Iris-virginica has a very low value, read it again, please. You have missed e13 in it. Your output is perfect.

Iris-virginica result is written in scientific notation.

[vorticalbox]

Iris-virginica has a very low value, read it again, please. You have missed e13 in it. Your output is perfect.

Thanks, I worked that out after I re-read your comment. I am guessing this is just a javascript thing with floats?

the result is actually

0.0000000000002462011443240403

[robertleeplummerjr]

It's exponent values. An effort to make it more computationally easy and/or visually friendly. 13 zeros vs 2 zeros. Happens to the best of us.

[vorticalbox]

It's exponent values. An effort to make it more computationally easy and/or visually friendly. 13 zeros vs 2 zeros. Happens to the best of us.

is there an option to stop it from doing this and just return the float?

[robertleeplummerjr]

Sadly it is a core part of javascript, but it _is_ a float, and is really just for displaying purposes only. I wouldn't worry about it.

[robertleeplummerjr]

The other option you can go with is formatting the output as a string, such as using .toFixed(14).

Example:

2.462011443240403e-13.toFixed(14); // -> "0.00000000000025"