Fuzzy-based feature and instance recovery

The severe class distribution shews the presence of underrepresented data, which has great effects on the performance of learning algorithm, is still a challenge of data mining and machine learning. Lots of researches currently focus on experimental comparison of the existing re-sampling approaches. We believe it requires new ways of constructing better algorithms to further balance and analyse the data set. This paper presents a Fuzzy-based Information Decomposition oversampling (FIDoS) algorithm used for handling the imbalanced data. Generally speaking, this is a new way of addressing imbalanced learning problems from missing data perspective. First, we assume that there are missing instances in the minority class that result in the imbalanced dataset. Then the proposed algorithm which takes advantages of fuzzy membership function is used to transfer information to the missing minority class instances. Finally, the experimental results demonstrate that the proposed algorithm is more practical and applicable compared to sampling techniques.

Statistical detection of online drifting twitter spam

Spam has become a critical problem in online social networks. This paper focuses on Twitter spam detection. Recent research works focus on applying machine learning techniques for Twitter spam detection, which make use of the statistical features of tweets. We observe existing machine learning based detection methods suffer from the problem of Twitter spam drift, i.e., the statistical properties of spam tweets vary over time. To avoid this problem, an effective solution is to train one twitter spam classifier every day. However, it faces a challenge of the small number of imbalanced training data because labelling spam samples is time-consuming. This paper proposes a new method to address this challenge. The new method employs two new techniques, fuzzy-based redistribution and asymmetric sampling. We develop a fuzzy-based information decomposition technique to re-distribute the spam class and generate more spam samples. Moreover, an asymmetric sampling technique is proposed to re-balance the sizes of spam samples and non-spam samples in the training data. Finally, we apply the ensemble technique to combine the spam classifiers over two different training sets. A number of experiments are performed on a real-world 10-day ground-truth dataset to evaluate the new method. Experiments results show that the new method can significantly improve the detection performance for drifting Twitter spam.

Frequency decomposition techniques for increased discriminative 3D facial information capture

Eigen-based techniques and other monolithic approaches to face recognition have long been a cornerstone in the face recognition community due to the high dimensionality of face images. Eigen-face techniques provide minimal reconstruction error and limit high-frequency content while linear discriminant-based techniques (fisher-faces) allow the construction of subspaces which preserve discriminatory information. This paper presents a frequency decomposition approach for improved face recognition performance utilising three well-known techniques: Wavelets; Gabor / Log-Gabor; and the Discrete Cosine Transform. Experimentation illustrates that frequency domain partitioning prior to dimensionality reduction increases the information available for classification and greatly increases face recognition performance for both eigen-face and fisher-face approaches.

Extracting evolutionary information from the spectral decomposition of early-type galaxies.

Holding the major share of stellar mass in galaxies and being also old and passively evolving, early-type galaxies (ETGs) are the primary probes in investigating these various evolution scenarios, as well as being useful means to provide insights on cosmological parameters. In this thesis work I focused specifically on ETGs and on their capability in constraining galaxy formation and evolution; in particular, the principal aims were to derive some of the ETGs evolutionary parameters, such as age, metallicity and star formation history (SFH) and to study their age-redshift and mass-age relations. In order to infer galaxy physical parameters, I used the public code STARLIGHT: this program provides a best fit to the observed spectrum from a combination of many theoretical models defined in user-made libraries. the comparison between the output and input light-weighted ages shows a good agreement starting from SNRs of ∼ 10, with a bias of ∼ 2.2% and a dispersion 3%. Furthermore, also metallicities and SFHs are well reproduced. In the second part of the thesis I performed an analysis on real data, starting from Sloan Digital Sky Survey (SDSS) spectra. I found that galaxies get older with cosmic time and with increasing mass (for a fixed redshift bin); absolute light-weighted ages, instead, result independent from the fitting parameters or the synthetic models used. Metallicities, instead, are very similar from each other and clearly consistent with the ones derived from the Lick indices. The predicted SFH indicates the presence of a double burst of star formation. Velocity dispersions and extinctiona are also well constrained, following the expected behaviours. As a further step, I also fitted single SDSS spectra (with SNR∼ 20), to verify that stacked spectra gave the same results without introducing any bias: this is an important check, if one wants to apply the method at higher z, where stacked spectra are necessary to increase the SNR. Our upcoming aim is to adopt this approach also on galaxy spectra obtained from higher redshift Surveys, such as BOSS (z ∼ 0.5), zCOSMOS (z 1), K20 (z ∼ 1), GMASS (z ∼ 1.5) and, eventually, Euclid (z 2). Indeed, I am currently carrying on a preliminary study to estabilish the applicability of the method to lower resolution, as well as higher redshift (z 2) spectra, just like the Euclid ones.

Composite load model decomposition : induction motor contribution

In this thesis, a new technique has been developed for determining the composition of a collection of loads including induction motors. The application would be to provide a representation of the dynamic electrical load of Brisbane so that the ability of the power system to survive a given fault can be predicted. Most of the work on load modelling to date has been on post disturbance analysis, not on continuous on-line models for loads. The post disturbance methods are unsuitable for load modelling where the aim is to determine the control action or a safety margin for a specific disturbance. This thesis is based on on-line load models. Dr. Tania Parveen considers 10 induction motors with different power ratings, inertia and torque damping constants to validate the approach, and their composite models are developed with different percentage contributions for each motor. This thesis also shows how measurements of a composite load respond to normal power system variations and this information can be used to continuously decompose the load continuously and to characterize regarding the load into different sizes and amounts of motor loads.

An aspect query language model based on query decomposition and high-order contextual term associations

In information retrieval (IR) research, more and more focus has been placed on optimizing a query language model by detecting and estimating the dependencies between the query and the observed terms occurring in the selected relevance feedback documents. In this paper, we propose a novel Aspect Language Modeling framework featuring term association acquisition, document segmentation, query decomposition, and an Aspect Model (AM) for parameter optimization. Through the proposed framework, we advance the theory and practice of applying high-order and context-sensitive term relationships to IR. We first decompose a query into subsets of query terms. Then we segment the relevance feedback documents into chunks using multiple sliding windows. Finally we discover the higher order term associations, that is, the terms in these chunks with high degree of association to the subsets of the query. In this process, we adopt an approach by combining the AM with the Association Rule (AR) mining. In our approach, the AM not only considers the subsets of a query as “hidden” states and estimates their prior distributions, but also evaluates the dependencies between the subsets of a query and the observed terms extracted from the chunks of feedback documents. The AR provides a reasonable initial estimation of the high-order term associations by discovering the associated rules from the document chunks. Experimental results on various TREC collections verify the effectiveness of our approach, which significantly outperforms a baseline language model and two state-of-the-art query language models namely the Relevance Model and the Information Flow model

Distribution feeder loads classification and decomposition

Load in distribution networks is normally measured at the 11kV supply points; little or no information is known about the type of customers and their contributions to the load. This paper proposes statistical methods to decompose an unknown distribution feeder load to its customer load sector/subsector profiles. The approach used in this paper should assist electricity suppliers in economic load management, strategic planning and future network reinforcements.

Chronological categorization and decomposition of customer loads

The majority of distribution utilities do not have accurate information on the constituents of their loads. This information is very useful in managing and planning the network, adequately and economically. Customer loads are normally categorized in three main sectors: 1) residential; 2) industrial; and 3) commercial. In this paper, penalized least-squares regression and Euclidean distance methods are developed for this application to identify and quantify the makeup of a feeder load with unknown sectors/subsectors. This process is done on a monthly basis to account for seasonal and other load changes. The error between the actual and estimated load profiles are used as a benchmark of accuracy. This approach has shown to be accurate in identifying customer types in unknown load profiles, and is used in cross-validation of the results and initial assumptions.

Finding additional semantic entity information for search engines

Entity-oriented search has become an essential component of modern search engines. It focuses on retrieving a list of entities or information about the specific entities instead of documents. In this paper, we study the problem of finding entity related information, referred to as attribute-value pairs, that play a significant role in searching target entities. We propose a novel decomposition framework combining reduced relations and the discriminative model, Conditional Random Field (CRF), for automatically finding entity-related attribute-value pairs from free text documents. This decomposition framework allows us to locate potential text fragments and identify the hidden semantics, in the form of attribute-value pairs for user queries. Empirical analysis shows that the decomposition framework outperforms pattern-based approaches due to its capability of effective integration of syntactic and semantic features.

Decomposition construction for secret sharing schemes with graph access structures in polynomial time

The purpose of this paper is to describe a new decomposition construction for perfect secret sharing schemes with graph access structures. The previous decomposition construction proposed by Stinson is a recursive method that uses small secret sharing schemes as building blocks in the construction of larger schemes. When the Stinson method is applied to the graph access structures, the number of such “small” schemes is typically exponential in the number of the participants, resulting in an exponential algorithm. Our method has the same flavor as the Stinson decomposition construction; however, the linear programming problem involved in the construction is formulated in such a way that the number of “small” schemes is polynomial in the size of the participants, which in turn gives rise to a polynomial time construction. We also show that if we apply the Stinson construction to the “small” schemes arising from our new construction, both have the same information rate.

On application of structural decomposition for process model abstraction

Real world business process models may consist of hundreds of elements and have sophisticated structure. Although there are tasks where such models are valuable and appreciated, in general complexity has a negative influence on model comprehension and analysis. Thus, means for managing the complexity of process models are needed. One approach is abstraction of business process models-creation of a process model which preserves the main features of the initial elaborate process model, but leaves out insignificant details. In this paper we study the structural aspects of process model abstraction and introduce an abstraction approach based on process structure trees (PST). The developed approach assures that the abstracted process model preserves the ordering constraints of the initial model. It surpasses pattern-based process model abstraction approaches, allowing to handle graph-structured process models of arbitrary structure. We also provide an evaluation of the proposed approach.

Efficient computation of causal behavioural profiles using structural decomposition

Identification of behavioural contradictions is an important aspect of software engineering, in particular for checking the consistency between a business process model used as system specification and a corresponding workflow model used as implementation. In this paper, we propose causal behavioural profiles as the basis for a consistency notion, which capture essential behavioural information, such as order, exclusiveness, and causality between pairs of activities. Existing notions of behavioural equivalence, such as bisimulation and trace equivalence, might also be applied as consistency notions. Still, they are exponential in computation. Our novel concept of causal behavioural profiles provides a weaker behavioural consistency notion that can be computed efficiently using structural decomposition techniques for sound free-choice workflow systems if unstructured net fragments are acyclic or can be traced back to S- or T-nets.