Collaborating Authors

Multiset Feature Learning for Highly Imbalanced Data Classification

AAAI Conferences

With the expansion of data, increasing imbalanced data has emerged. When the imbalance ratio of data is high, most existing imbalanced learning methods decline in classification performance. To address this problem, a few highly imbalanced learning methods have been presented. However, most of them are still sensitive to the high imbalance ratio. This work aims to provide an effective solution for the highly imbalanced data classification problem. We conduct highly imbalanced learning from the perspective of feature learning. We partition the majority class into multiple blocks with each being balanced to the minority class and combine each block with the minority class to construct a balanced sample set. Multiset feature learning (MFL) is performed on these sets to learn discriminant features. We thus propose an uncorrelated cost-sensitive multiset learning (UCML) approach. UCML provides a multiple sets construction strategy, incorporates the cost-sensitive factor into MFL, and designs a weighted uncorrelated constraint to remove the correlation among multiset features. Experiments on five highly imbalanced datasets indicate that: UCML outperforms state-of-the-art imbalanced learning methods.

New Liftable Classes for First-Order Probabilistic Inference

Neural Information Processing Systems

Statistical relational models provide compact encodings of probabilistic dependencies in relational domains, but result in highly intractable graphical models. The goal of lifted inference is to carry out probabilistic inference without needing to reason about each individual separately, by instead treating exchangeable, undistinguished objects as a whole. In this paper, we study the domain recursion inference rule, which, despite its central role in early theoretical results on domain-lifted inference, has later been believed redundant. We show that this rule is more powerful than expected, and in fact significantly extends the range of models for which lifted inference runs in time polynomial in the number of individuals in the domain. This includes an open problem called S4, the symmetric transitivity model, and a first-order logic encoding of the birthday paradox.

Bounded Programs: A New Decidable Class of Logic Programs with Function Symbols

AAAI Conferences

While function symbols are widely acknowledged as an important feature in logic programming, they make common inference tasks undecidable. To cope with this problem, recent research has focused on identifying classes of logic programs imposing restrictions on the use of function symbols, but guaranteeing decidability of common inference tasks. This has led to several criteria, called termination criteria, providing sufficient conditions for a program to have finitely many stable models, each of finite size.This paper introduces the new class of bounded programs which guarantees the aforementioned property and strictly includes the classes of programs determined by current termination criteria. Different results on the correctness, the expressiveness, and the complexity of the class of bounded programs are presented.

This $16 online math class is great for home schooling — or for yourself


TL;DR: Get help understanding sets, logic, stats, and more with the Discrete Mathematics Course for $15.99, a 91% savings as of Aug. 17. Remember that joke when you were in school about math being useless? Let's face it: You were wrong. And while it's highly unlikely that you'll have to SIN COS TAN your way through life, you still need to master how numbers work, especially if you or your newly-homeschooled child is gunning for a highly technical job in STEM. If you're diving deep into the world of mathematics and computer science or are trying to find some additional learning materials for your kids, the Discrete Mathematics course is a great option.

Discovering Classes of Strongly Equivalent Logic Programs Artificial Intelligence

In this paper we apply computer-aided theorem discovery technique to discover theorems about strongly equivalent logic programs under the answer set semantics. Our discovered theorems capture new classes of strongly equivalent logic programs that can lead to new program simplification rules that preserve strong equivalence. Specifically, with the help of computers, we discovered exact conditions that capture the strong equivalence between a rule and the empty set, between two rules, between two rules and one of the two rules, between two rules and another rule, and between three rules and two of the three rules.