Published
Edited
Aug 2, 2020
5 stars
Autoencoder = require('https://bundle.run/autoencoder@0.0.2')
data = require("@observablehq/iris").then(l =>
l.map(d =>
Object.assign([d.sepalLength, d.sepalWidth, d.petalLength, d.petalWidth], {
color: classify(d.species)
})
)
)
classify = d3.scaleOrdinal(d3.schemeAccent)
ae = {
replay;
const ae = new Autoencoder({
nInputs: data[0].length,
nHidden: 4,
scale: true,
activation
});
ae.fit(data, {
batchSize: 50,
iterations: 2000,
method,
stepSize: 0.05
});
return ae;
}
Y = ae.encode(data)
ae.decode(Y)
ae.predict(data) // same as decode(encode(data))
distances = ae
.predict(data)
.map((predicted, i) => Math.hypot(...predicted.map((d, j) => d - data[i][j])))
score = d3.quantile(distances, .9)
d3 = require("d3@6.0.0-rc.2")
import { select } from "@jashkenas/inputs"
Purpose-built for displays of data
Observable is your go-to platform for exploring data and creating expressive data visualizations. Use reactive JavaScript notebooks for prototyping and a collaborative canvas for visual data exploration and dashboard creation.
