import { RegressorBase } from '../base';
import { Params } from '../base/estimator';
export interface LinearSVRProps {
    epsilon?: number;
    C?: number;
    maxIter?: number;
    /** @deprecated ignored — Pegasos uses the schedule eta_t = 1/(lambda*t) */
    learningRate?: number;
    tol?: number;
    randomState?: number;
}
/**
 * Linear SVR trained with Pegasos-style stochastic subgradient descent on
 * lambda/2 ||w||^2 + (1/n) sum max(0, |w.x + b - y| - epsilon), where
 * lambda = 1/(n*C) so the C semantics match sklearn's LinearSVR.
 */
export declare class LinearSVR extends RegressorBase {
    private epsilon;
    private C;
    private maxIter;
    /** @deprecated kept only so params round-trip; the Pegasos schedule ignores it */
    private learningRate?;
    private tol;
    private randomState?;
    private weights;
    private bias;
    private fitted;
    constructor(props?: LinearSVRProps);
    getParams(): Params;
    private objective;
    fit(trainX: number[][], trainY: number[]): void;
    predict(testX: number[][]): number[];
    get coef(): number[];
}