Gabor filters are widely utilized to detect iris texture information in several state-of-the-art recognition systems. However, the proper kernels and generative pattern of features need be predetermined application. The traditional empirical shallow encoding ways incapable dealing with such complex variations imaging including illumination, aging, deformation, device variations. Thereby, an adaptive filter selection strategy deep learning architecture presented. We first employ particle...

Although words are basic semantic units in text, phrases, and expressions contain additional information, which is important for text classification. To capture this traditional algorithms extract composite features via word sequences or co-occurrences, such as bigrams termsets, but ignore the influence of stop punctuation, results huge amounts weak features. In paper, we propose a structure-based algorithm to Termsets that cross punctuation marks excluded. eliminate redundancy, novel...

Data with categorical attributes are ubiquitous in the real world. However, existing partitional clustering algorithms for data prone to fall into local optima. To address this issue, paper we propose a novel algorithm, ABC-K-Modes (Artificial Bee Colony based on K-Modes), traditional k-modes algorithm and artificial bee colony approach. In our approach, first introduce one-step procedure, then integrate procedure approach deal data. search process performed by scout bees, adopt multi-source...

Clustering analysis, as an important technique in data mining, aims to identify the nature groups or clusters of objects attribute space. Data real-world applications are commonly described by both numeric and categorical attributes. In this research, considering that partitional clustering algorithms designed for type mixed prone get trapped into local optima cuckoo search approach is efficient solving global optimization problems, we propose CCS-K-Prototypes, a novel algorithm based on...

Most of the initialization approaches are dedicated to partitional clustering algorithms which process categorical or numerical data only. However, in real-world applications, objects with both numeric and features ubiquitous. The coexistence attributes make methods designed for single-type inapplicable mixed-type data. Furthermore, best our knowledge, existing data, initial cluster centers determined randomly. In this paper, we propose a novel method mixed clustering. proposed method,...

Clustering data with both numeric and categorical attributes is of great importance as such are ubiquitous in real-world problems. Multi-view learning approaches have proven to be more effective having better generalisation ability compared single-view many However, most the existing clustering algorithms developed for mixed single-view. In this research, we propose a novel multi-view algorithm based on k-prototypes (which term K-Prototypes) data. To best our knowledge, proposed K-Prototypes...

Acetylation on lysine is a widespread post-translational modification which reversible and plays crucial role in some biological activities. To better understand the mechanism, it necessary to identify acetylation sites proteins accurately. Computational methods are popular because they more convenient faster than experimental methods. In this study, we proposed new computational method predict human by combining sequence features structural including physicochemical property (PCP), position...

Community detection is important for many complex network applications. A major challenge lies in that the number of communities a given social usually unknown. This paper presents new community algorithm-Distance Centrality based Detection (DCCD). The proposed method capable detecting without preset number. has two components. First we choose initial center nodes by calculating centrality each node using their distance information. Then measure similarity between and other network, assign...

Data objects with both numeric and categorical attributes are prevalent in many real-world applications. However, most of the partitional clustering algorithms dealing such data may trap into local optima. To further promote performance, a novel algorithm, called ABC-K-Prototype s (Artificial Bee Colony based on K-Prototypes), is presented basis K-Prototypes search strategy artificial bee colony, chaos theory. In approach, one-step k-prototypes procedure first given, then this combined...

This paper is devoted to the study of periodic solutions first-order vector differential equations $x'(t)+f(t,x(t))=0$. We first introduce concepts positive, negative and mixed-type solutions. Then, by using a fixed point theorem in cones, we obtain some existence multiplicity results such Furthermore, also present examples illustrate our main results.