Many different metrics are used in machine learning and data mining to build and evaluate models. However, there is no general theory of machine learning metrics, that could answer questions such as: When we simultaneously want to optimise two criteria, how can or should they be traded off? Some metrics are inherently independent of class and misclassification cost distributions, while other ar...

In this paper, we present the case for a declarative foundation for data-intensive machine learning systems. Instead of creating a new system for each specific flavor of machine learning task, or hardcoding new optimizations, we argue for the use of recursive queries to program a variety of machine learning systems. By taking this approach, database query optimization techniques can be utilized...

Classification is a learning function that maps a given data item into one of several predefined classes. It is a data analysis technique that extracts models describing important data classes and predicts future values. Basically, classification techniques have better capability to handle a wider variety of datasets than regression. It can also be described as a supervised learning algorithm i...

" Motivation Limited Domain relations (LD-relations) " What happens for out of domain arguments? Difficulty parsing complex mathematical specifications " What does that do, again? " Objective Clear, formal representation of requirements specifications in a tabular format " Mathematic foundation " Possible for both human and machine parsing " Better partitioned specifications " Techniques Used M...

In this paper, a co-axial dual-mechanical ports flux-switching permanent magnet (CADMP-FSPM) machine for hybrid electric vehicles (HEVs) is proposed and investigated, which is comprised of two conventional co-axial FSPM machines, namely one high-speed inner rotor machine and one low-speed outer rotor machine and a non-magnetic ring sandwiched in between. Firstly, the topology and operation prin...

Scholar metadata has traditionally centered on descriptive representation, which has been used as a foundation for scholarly publication repositories and academic information retrieval systems. In this paper we propose innovative and economic methods of generating knowledge-­‐based structural metadata (Structural Keywords) using a combination of NLP-­‐based machine learning techniques and human...

Y. Huang acknowledges partial support from a UCF Graduate College Presidential Fellowship and National Science Foundation (NS F) grant No. 1200566. C. Li acknowledges partial support from NSF grants No. 0806931 and No. 0963146. M. Georgiopoulos acknowledges partial support from NSF grants No. 1161228 and No. 0525429. G. G. Anagnostopoulos acknowledges partial support from NSF grant No. 1263011....

In this paper we explore Machine Comprehension, a subset of Natural Language Processing, using data from SQuAD. We take an iterative approach to designing our prediction model, starting with a very basic foundation, extending it and replacing components piece by piece while testing for effectiveness. Our model achieves a 62.3 F1 score on the SQuAD test set.

When trained on very large parallel corpora, the phrase table component of a machine translation system grows to consume vast computational resources. In this paper, we introduce a novel pruning criterion that places phrase table pruning on a sound theoretical foundation. Systematic experiments on four language pairs under various data conditions show that our principled approach is superior to...

The use of wind turbines to generate electricity has increased in recent years. One of the most important parts of a wind turbine is the foundation, which should be designed accurately because it is influenced by difference forces. Soil cannot carry tension stress; thus, when a wind turbine foundation is applied eccentricity forces, a gap appears between the soil and foundation. The gap will ha...