holisticai.bias.mitigation.BlindSpotAwareMF

class holisticai.bias.mitigation.BlindSpotAwareMF(*args, **kargs)

Blind Spot Aware Matrix Factorization

A blind spot aware Matrix Factorization takes into account the blind spot inherent in the learning phase of the RS. The blind spot size is the number of item with a predicted ratings Ru,i that is smaller than a threshold.

Parameters

Kint

Specifies the number of dimensions.

betafloat

Parameter used to update P and Q.

stepsint

Number of iterations.

alphafloat

Model parameter. Alpha is the learning rate.

lamdafloat

Model parameter. Lambda is the regularization parameter.

verboseint

If >0, will show progress percentage.

Examples

>>> from holisticai.bias.mitigation import BlindSpotAwareMF
>>> mitigator = BlindSpotAwareMF(**params)
>>> mitigator.fit(data_matrix)

References

[1]

Sun, Wenlong, et al. “Debiasing the human-recommender system feedback loop in collaborative filtering.” Companion Proceedings of The 2019 World Wide Web Conference. 2019.

fit(X: ndarray, **kargs)

Fit model

Parameters

Xmatrix-like

rating matrix, 0 means non-raked cases.

P0matrix-like (optional)

Initial P matrix (numUsers, K)

Q0matrix-like (optional)

Initial P matrix (numItems, K)

Returns

self