Examples with structures._RankItem structures._RankItem used on opensource projects

Example 16 with structures._RankItem

use of structures._RankItem in project IR_Base by Linda-sunshine.

the class CoLinAdaptWithDiffFeatureGroups method calculateFuncValue.

// There is still issue in calculating R2 since we don't know which set to use for a user.
@Override
protected double calculateFuncValue(_AdaptStruct u) {
    double fValue = super.calculateFuncValue(u), R1 = 0, R2 = 0, diffA, diffB;
    _CoLinAdaptDiffFvGroupsStruct ui = (_CoLinAdaptDiffFvGroupsStruct) u, uj;
    // Add R1 for another class.
    for (int i = 0; i < m_dimB; i++) {
        // (a[i]-1)^2
        R1 += m_eta1 * (ui.getScalingB(i) - 1) * (ui.getScalingB(i) - 1);
        // b[i]^2
        R1 += m_eta2 * ui.getShiftingB(i) * ui.getShiftingB(i);
    }
    // R2 regularization
    for (_RankItem nit : ui.getNeighbors()) {
        uj = (_CoLinAdaptDiffFvGroupsStruct) m_userList.get(nit.m_index);
        diffA = 0;
        diffB = 0;
        // We also need to sum over the other set of parameters.
        for (int k = 0; k < m_dimB; k++) {
            diffA += (ui.getScalingB(k) - uj.getScalingB(k)) * (ui.getScalingB(k) - uj.getScalingB(k));
            diffB += (ui.getShiftingB(k) - uj.getShiftingB(k)) * (ui.getShiftingB(k) - uj.getShiftingB(k));
        }
        R2 += nit.m_value * (m_eta3 * diffA + m_eta4 * diffB);
    }
    return fValue + R1 + R2;
}

Example 17 with structures._RankItem

use of structures._RankItem in project IR_Base by Linda-sunshine.

the class CoLinAdaptWithNeighborhoodLearning method constructXijs.

// Construct all training instances.
public void constructXijs() {
    // Index of neighbors.
    int j = 0;
    _CoLinAdaptStruct ui;
    m_xijs = new double[m_topK * m_userList.size()][];
    for (int i = 0; i < m_userList.size(); i++) {
        ui = (_CoLinAdaptStruct) m_userList.get(i);
        j = 0;
        // Traverse all neighbors.
        for (_RankItem nit : ui.getNeighbors()) {
            // Construct the training instance.
            m_xijs[m_topK * i + j] = constructXij(ui.getUser(), m_userList.get(nit.m_index).getUser());
            j++;
        }
    }
}

Example 18 with structures._RankItem

use of structures._RankItem in project IR_Base by Linda-sunshine.

the class MTCoLinAdapt method calculateFuncValue.

@Override
protected // Calculate the function value of the new added instance.
double calculateFuncValue(_AdaptStruct u) {
    double fValue = super.calculateFuncValue(u), R2 = 0, diffA, diffB;
    // R2 regularization
    _CoLinAdaptStruct ui = (_CoLinAdaptStruct) u, uj;
    for (_RankItem nit : ui.getNeighbors()) {
        uj = (_CoLinAdaptStruct) m_userList.get(nit.m_index);
        diffA = 0;
        diffB = 0;
        for (int k = 0; k < m_dim; k++) {
            diffA += (ui.getScaling(k) - uj.getScaling(k)) * (ui.getScaling(k) - uj.getScaling(k));
            diffB += (ui.getShifting(k) - uj.getShifting(k)) * (ui.getShifting(k) - uj.getShifting(k));
        }
        R2 += nit.m_value * (m_eta3 * diffA + m_eta4 * diffB);
    }
    return fValue + R2;
}

Example 19 with structures._RankItem

use of structures._RankItem in project IR_Base by Linda-sunshine.

the class WeightedAvgAdapt method setPersonalizedModel.

@Override
public void setPersonalizedModel() {
    _CoLinAdaptStruct ui;
    double[] pWeights;
    int offset;
    for (int i = 0; i < m_userList.size(); i++) {
        ui = (_CoLinAdaptStruct) m_userList.get(i);
        pWeights = ui.getPWeights();
        offset = i * m_dim;
        for (int n = 0; n < m_dim; n++) pWeights[n] = ui.getSelfSim() * _CoLinAdaptStruct.sharedA[offset + n];
        // traverse all the neighbors
        for (_RankItem nit : ui.getNeighbors()) {
            // get neighbor's index
            offset = nit.m_index * m_dim;
            for (int n = 0; n < m_dim; n++) pWeights[n] += nit.m_value * _CoLinAdaptStruct.sharedA[offset + n];
        }
    }
}

Example 20 with structures._RankItem

use of structures._RankItem in project IR_Base by Linda-sunshine.

the class WeightedAvgTransAdapt method gradientByFunc.

// shared gradient calculation by batch and online updating
@Override
protected void gradientByFunc(_AdaptStruct u, _Doc review, double weight) {
    _CoLinAdaptStruct ui = (_CoLinAdaptStruct) u;
    int n, k, offsetj;
    // general enough to accommodate both LinAdapt and CoLinAdapt
    int offset = m_dim * ui.getId();
    double delta = weight * (review.getYLabel() - logit(review.getSparse(), ui));
    if (m_LNormFlag)
        delta /= getAdaptationSize(ui);
    // Current user's info: Bias term + other features.
    // \theta_{ii}*w_g[0]*x_0 and x_0=1
    m_g[offset] -= delta * ui.getSelfSim() * m_gWeights[0];
    // \theta_{ii}*x_0
    m_g[offset + m_dim] -= delta * ui.getSelfSim();
    for (_SparseFeature fv : review.getSparse()) {
        n = fv.getIndex() + 1;
        k = m_featureGroupMap[n];
        // \theta_{ii}*x_d
        m_g[offset + k] -= delta * ui.getSelfSim() * m_gWeights[n] * fv.getValue();
        m_g[offset + k + m_dim] -= delta * ui.getSelfSim() * fv.getValue();
    }
    // Neighbors' info: Bias term + other features.
    for (_RankItem nit : ui.getNeighbors()) {
        offsetj = 2 * m_dim * nit.m_index;
        // Bias term.
        // neighbors' bias term.
        m_g[offsetj] -= delta * nit.m_value * m_gWeights[0];
        m_g[offsetj + m_dim] -= delta * nit.m_value;
        for (_SparseFeature fv : review.getSparse()) {
            n = fv.getIndex() + 1;
            k = m_featureGroupMap[n];
            // neighbors' other features.
            m_g[offsetj + k] -= delta * nit.m_value * m_gWeights[n] * fv.getValue();
            m_g[offsetj + m_dim + k] -= delta * nit.m_value * fv.getValue();
        }
    }
}