Scalable Training of Markov Logic Networks Using Approximate Counting
Scalable Training of Markov Logic Networks Using Approximate Counting
复制标题
使用近似计数的马尔可夫逻辑网络的可扩展训练
DOI：
10.1609/aaai.v30i1.10119
复制
发表时间：
2016
期刊：
影响因子：
--
通讯作者：
Vibhav Gogate
中科院分区：
文献类型：
--
作者：
Somdeb Sarkhel;D. Venugopal;T. Pham;Parag Singla;Vibhav Gogate
在本文中，我们提出了马尔可夫逻辑网络的原则权重学习算法，该算法可以很容易地扩展到比现有算法更大的数据集和应用领域。我们方法的主要思想是使用近似计数技术来大幅降低权重学习中计算最密集的子步骤的复杂性：计算一阶公式的基数，该一阶公式在给定所有随机变量的真值分配的情况下为真。我们推导了新算法性能的理论界限，并通过实验证明，它们的速度要快几个数量级，并且达到与现有方法相同或更好的精度。
In this paper, we propose principled weight learning algorithms for Markov logic networks that can easily scale to much larger datasets and application domains than existing algorithms. The main idea in our approach is to use approximate counting techniques to substantially reduce the complexity of the most computation intensive sub-step in weight learning: computing the number of groundings of a first-order formula that evaluate to true given a truth assignment to all the random variables. We derive theoretical bounds on the performance of our new algorithms and demonstrate experimentally that they are orders of magnitude faster and achieve the same accuracy or better than existing approaches.