Multi-task learning using bayesian model with enforced sparsity and leveraging of task correlations
View Patent ↗Multi-task regression or classification includes optimizing parameters of a Bayesian model representing relationships between D features and P tasks, where D≧1 and P≧1, respective to training data comprising sets of values for the D features annotated with values for the P tasks. The Bayesian model includes a matrix-variate prior having features and tasks dimensions of dimensionality D and P respectively. The matrix-variate prior is partitioned into a plurality of blocks, and the optimizing of parameters of the Bayesian model includes inferring prior distributions for the blocks of the matrix-variate prior that induce sparseness of the plurality of blocks. Values of the P tasks are predicted for a set of input values for the D features using the optimized Bayesian model. The optimizing also includes decomposing the matrix-variate prior into a product of matrices including a matrix of reduced rank in the tasks dimension that encodes correlations between tasks.
1. An apparatus comprising:
an electronic data processing device configured to perform a method comprising:
constructing a Bayesian model representing relationships between a plurality of features and a plurality of tasks wherein the Bayesian model includes a matrix-variate Gaussian scaled mixture prior having a features dimension and a tasks dimension and wherein the matrix-variate Gaussian scaled mixture prior is partitioned into a plurality of blocks;
generating an optimized Bayesian model by optimizing parameters of the Bayesian model respective to training data comprising sets of feature values annotated with values for tasks of the plurality of tasks wherein the optimizing includes inferring prior distributions for the blocks of the matrix-variate Gaussian scaled mixture prior that induce sparseness of the plurality of blocks; and
predicting values of tasks of the plurality of tasks for an input set of feature values using the optimized Bayesian model.
2. The apparatus of claim 1 , wherein the optimizing comprises:
decomposing the matrix-variate Gaussian scaled mixture prior into a product of matrices including a matrix of reduced rank in the tasks dimension that encodes positive or negative correlations between tasks of the plurality of tasks.
3. The apparatus of claim 1 , wherein the predicting does not use a feature value of the input set of feature values for predicting the value of a task when the expectation of the element of the matrix-variate Gaussian scaled mixture prior corresponding to the feature-task pair has a zero value.
4. An apparatus comprising:
an electronic data processing device configured to perform a method comprising:
constructing a Bayesian model representing relationships between a plurality of features and a plurality of tasks wherein the Bayesian model includes a matrix-variate prior having a features dimension and a tasks dimension and wherein the matrix-variate prior is partitioned into a plurality of blocks;
generating an optimized Bayesian model by optimizing parameters of the Bayesian model respective to training data comprising sets of feature values annotated with values for tasks of the plurality of tasks wherein the optimizing includes inferring prior distributions for the blocks of the matrix-variate prior that induce sparseness of the plurality of blocks; and
predicting values of tasks of the plurality of tasks for an input set of feature values using the optimized Bayesian model, wherein the predicting does not use a feature value of the input set of feature values for predicting the value of a task when the expectation of the element of the matrix-variate prior corresponding to the feature-task pair has a zero value.
5. An apparatus comprising:
an electronic data processing device configured to perform a method comprising:
constructing a Bayesian model representing relationships between a plurality of features and a plurality of tasks wherein the Bayesian model includes a matrix-variate prior having a features dimension and a tasks dimension and wherein the matrix-variate prior is partitioned into a plurality of blocks;
generating an optimized Bayesian model by optimizing parameters of the Bayesian model respective to training data comprising sets of feature values annotated with values for tasks of the plurality of tasks wherein the optimizing employs a variational expectation-maximization (variational E-M) algorithm and wherein the optimizing includes inferring prior distributions for the blocks of the matrix-variate prior that induce sparseness of the plurality of blocks; and
predicting values of tasks of the plurality of tasks for an input set of feature values using the optimized Bayesian model.
6. A method comprising:
generating an optimized Bayesian model by optimizing parameters of a Bayesian model representing relationships between a plurality of features and a plurality of tasks respective to training data comprising sets of feature values annotated with values for tasks of the plurality of tasks, the Bayesian model including a matrix-variate Gaussian scaled mixture prior having a features dimension and a tasks dimension, the optimizing including decomposing the matrix-variate Gaussian scaled mixture prior into a product of matrices including a matrix of reduced rank in the tasks dimension that encodes positive or negative correlations between tasks of the plurality of tasks; and
predicting values of tasks of the plurality of tasks for an input set of feature values using the optimized Bayesian model;
wherein the generating and predicting are performed by an electronic data processing device.
7. The method of claim 6 , wherein the matrix-variate Gaussian scaled mixture prior is partitioned into a plurality of blocks and the optimizing of parameters of the Bayesian model includes inferring prior distributions for the blocks of the matrix-variate Gaussian scaled mixture prior that induce sparseness of the plurality of blocks.
8. The method of claim 7 , wherein the predicting does not use a feature value of the input set of feature values for predicting the value of a task when the expectation of the element of the matrix-variate Gaussian scaled mixture prior corresponding to the feature-task pair has a zero value.
9. The method of claim 6 , wherein the optimizing employs a variational expectation-maximization (variational E-M) algorithm.
10. A non-transitory storage medium storing instructions executable by an electronic data processing device to perform operations including (1) generating an optimized Bayesian model by optimizing parameters of a Bayesian model representing relationships between D features and P tasks, where D≧1 and P≧1, respective to training data comprising sets of values for the D features annotated with values for the P tasks, the Bayesian model including a matrix-variate prior Wε P×D or Wε D×P having a features dimension of dimensionality D corresponding to the D features and a tasks dimension of dimensionality P corresponding to the P tasks, wherein the matrix-variate prior is partitioned into a plurality of blocks W i ε P×D or W i ε D i ×P where i=1, . . . , Q and Σ i=1 Q D i =D and the optimizing of parameters of the Bayesian model includes inferring prior distributions for the blocks of the matrix-variate prior that induce sparseness of the plurality of blocks and (2) predicting values of the P tasks for a set of input values for the D features using the optimized Bayesian model.
11. The non-transitory storage medium of claim 10 , wherein the optimizing comprises decomposing the matrix-variate prior into a product of matrices including a matrix of reduced rank in the tasks dimension that encodes positive or negative correlations between tasks of the P tasks.
12. The non-transitory storage medium of claim 11 , wherein the matrix-variate prior comprises a matrix-variate Gaussian scaled mixture prior.
13. The non-transitory storage medium of claim 11 , wherein the predicting does not use the input value for a d th feature for predicting the value of a p th task when the element of the matrix-variate prior at the d th feature dimension and p th task dimension has a zero value.
14. The non-transitory storage medium of claim 11 , wherein the optimizing employs a variational expectation-maximization (variational E-M) algorithm.
15. The non-transitory storage medium of claim 10 , wherein the predicting does not use the input value for a d th feature for predicting the value of a p th task when the expectation of the element of the matrix-variate prior at the d th feature dimension and p th task dimension has a zero value.
16. The non-transitory storage medium of claim 10 , wherein the matrix-variate prior comprises a matrix-variate Gaussian scaled mixture prior.