// FIXME: Use weights
double initial_MSE(Vec train, Vec test) {
if (train.isEnum()) {
// Guess the class of the most populous class; call the fraction of those
// Q. Then Q of them are "mostly correct" - error is (1-Q) per element.
// The remaining 1-Q elements are "mostly wrong", error is Q (our guess,
// which is wrong).
int cls = ArrayUtils.maxIndex(train.bins());
double guess = train.bins()[cls] / (train.length() - train.naCnt());
double actual = test.bins()[cls] / (test.length() - test.naCnt());
return guess * guess + actual - 2.0 * actual * guess;
} else { // Regression
// Guessing the training data mean, but actual is validation set mean
double stddev = test.sigma();
double bias = train.mean() - test.mean();
return stddev * stddev + bias * bias;
}
}
// Pick most common cat level for each cluster_centers' cat columns
private static double[][] max_cats(double[][] centers, long[][][] cats, String[][] isCats) {
for (int clu = 0; clu < centers.length; clu++)
for (int col = 0; col < centers[0].length; col++)
if (isCats[col] != null) centers[clu][col] = ArrayUtils.maxIndex(cats[clu][col]);
return centers;
}
// public final double[] rproxgrad_x(double[] u, double alpha) { return rproxgrad(u, alpha,
// _gamma_x, _regularization_x, RandomUtils.getRNG(_seed)); }
// public final double[] rproxgrad_y(double[] u, double alpha) { return rproxgrad(u, alpha,
// _gamma_y, _regularization_y, RandomUtils.getRNG(_seed)); }
static double[] rproxgrad(
double[] u, double alpha, double gamma, Regularizer regularization, Random rand) {
if (u == null || alpha == 0 || gamma == 0) return u;
double[] v = new double[u.length];
switch (regularization) {
case None:
return u;
case Quadratic:
for (int i = 0; i < u.length; i++) v[i] = u[i] / (1 + 2 * alpha * gamma);
return v;
case L2:
// Proof uses Moreau decomposition; see section 6.5.1 of Parikh and Boyd
// https://web.stanford.edu/~boyd/papers/pdf/prox_algs.pdf
double weight = 1 - alpha * gamma / ArrayUtils.l2norm(u);
if (weight < 0) return v; // Zero vector
for (int i = 0; i < u.length; i++) v[i] = weight * u[i];
return v;
case L1:
for (int i = 0; i < u.length; i++)
v[i] = Math.max(u[i] - alpha * gamma, 0) + Math.min(u[i] + alpha * gamma, 0);
return v;
case NonNegative:
for (int i = 0; i < u.length; i++) v[i] = Math.max(u[i], 0);
return v;
case OneSparse:
int idx = ArrayUtils.maxIndex(u, rand);
v[idx] = u[idx] > 0 ? u[idx] : 1e-6;
return v;
case UnitOneSparse:
idx = ArrayUtils.maxIndex(u, rand);
v[idx] = 1;
return v;
case Simplex:
// Proximal gradient algorithm by Chen and Ye in http://arxiv.org/pdf/1101.6081v2.pdf
// 1) Sort input vector u in ascending order: u[1] <= ... <= u[n]
int n = u.length;
int[] idxs = new int[n];
for (int i = 0; i < n; i++) idxs[i] = i;
ArrayUtils.sort(idxs, u);
// 2) Calculate cumulative sum of u in descending order
// cumsum(u) = (..., u[n-2]+u[n-1]+u[n], u[n-1]+u[n], u[n])
double[] ucsum = new double[n];
ucsum[n - 1] = u[idxs[n - 1]];
for (int i = n - 2; i >= 0; i--) ucsum[i] = ucsum[i + 1] + u[idxs[i]];
// 3) Let t_i = (\sum_{j=i+1}^n u[j] - 1)/(n - i)
// For i = n-1,...,1, set optimal t* to first t_i >= u[i]
double t = (ucsum[0] - 1) / n; // Default t* = (\sum_{j=1}^n u[j] - 1)/n
for (int i = n - 1; i >= 1; i--) {
double tmp = (ucsum[i] - 1) / (n - i);
if (tmp >= u[idxs[i - 1]]) {
t = tmp;
break;
}
}
// 4) Return max(u - t*, 0) as projection of u onto simplex
double[] x = new double[u.length];
for (int i = 0; i < u.length; i++) x[i] = Math.max(u[i] - t, 0);
return x;
default:
throw new RuntimeException("Unknown regularization function " + regularization);
}
}