@ai-on-browser/data-analysis-models/lib/model/hopfield.js

Data analysis model package without any dependencies

/**
 * Hopfield network
 */
export default class HopfieldNetwork {
	// https://www.tcom242242.net/entry/ai-2/deeplearning/%E3%83%8B%E3%83%A5%E3%83%BC%E3%83%A9%E3%83%AB%E3%83%8D%E3%83%83%E3%83%88/hopfieldnetwork/
	_normalize(x) {
		let max = -Infinity
		let min = Infinity
		for (let k = 0; k < x.length; k++) {
			for (let i = 0; i < x[k].length; i++) {
				max = Math.max(max, x[k][i])
				min = Math.min(min, x[k][i])
			}
		}
		for (let k = 0; k < x.length; k++) {
			for (let i = 0; i < x[k].length; i++) {
				x[k][i] = x[k][i] < (max + min) / 2 ? -1 : 1
			}
		}
	}

	/**
	 * Fit model.
	 * @param {Array<Array<number>>} x Training data
	 */
	fit(x) {
		this._normalize(x)
		this._w = []
		for (let i = 0; i < x[0].length; this._w[i++] = []);
		for (let k = 0; k < x.length; k++) {
			for (let i = 0; i < x[k].length; i++) {
				this._w[i][i] = 0
				for (let j = 0; j < i; j++) {
					this._w[i][j] = this._w[j][i] = x[k][i] * x[k][j]
				}
			}
		}
	}

	/**
	 * Return a energy value of the data.
	 * @param {Array<number>} x Sample data
	 * @returns {number} Energy value
	 */
	energy(x) {
		this._normalize([x])
		let e = 0
		for (let i = 0; i < this._w.length; i++) {
			for (let j = 0; j < this._w[i].length; j++) {
				e -= (this._w[i][j] * x[i] * x[j]) / 2
			}
		}
		return e
	}

	/**
	 * Returns predicted values.
	 * @param {Array<number>} x Sample data
	 * @returns {number[]} Predicted values
	 */
	predict(x) {
		this._normalize([x])
		const y = []
		for (let i = 0; i < x.length; i++) {
			y[i] = 0
			for (let j = 0; j < x.length; j++) {
				y[i] += this._w[i][j] * x[j]
			}
		}
		return y.map((v, i) => (v < 0 ? -1 : v > 0 ? 1 : x[i]))
	}
}