Domain Knowledge Uncertainty and Probabilistic Parameter Constraints
This work addresses the challenge of integrating human-provided domain knowledge with inherent uncertainty into generative and conditional models, offering a more robust alternative to hard constraints.
The paper tackles the problem of incorporating uncertain domain knowledge into machine learning models by proposing probabilistic constraints over parameters, which improves modeling accuracy when domain knowledge is inaccurate.
Incorporating domain knowledge into the modeling process is an effective way to improve learning accuracy. However, as it is provided by humans, domain knowledge can only be specified with some degree of uncertainty. We propose to explicitly model such uncertainty through probabilistic constraints over the parameter space. In contrast to hard parameter constraints, our approach is effective also when the domain knowledge is inaccurate and generally results in superior modeling accuracy. We focus on generative and conditional modeling where the parameters are assigned a Dirichlet or Gaussian prior and demonstrate the framework with experiments on both synthetic and real-world data.