Improved GART neural network model for pattern classification and rule extraction with application to power systems

Generalized adaptive resonance theory (GART) is a neural network model that is capable of online learning and is effective in tackling pattern classification tasks. In this paper, we propose an improved GART model (IGART), and demonstrate its applicability to power systems. IGART enhances the dynamics of GART in several aspects, which include the use of the Laplacian likelihood function, a new vigilance function, a new match-tracking mechanism, an ordering algorithm for determining the sequence of training data, and a rule extraction capability to elicit if-then rules from the network. To assess the effectiveness of IGART and to compare its performances with those from other methods, three datasets that are related to power systems are employed. The experimental results demonstrate the usefulness of IGART with the rule extraction capability in undertaking classification problems in power systems engineering.

Rule pruning in a fuzzy rule-based classification system

In this paper, we purpose a rule pruning strategy to reduce the number of rules in a fuzzy rule-based classification system.A confidence factor, which is formulated based on the compatibility of the rules with the input patterns is under deployed for rule pruning.The pruning strategy aims at reducing the complexity of the fuzzy classification system and, at the same time, maintaining the accuracy rate at a good level.To evaluate the effectiveness of the pruning strategy, two benchmark data sets are first tested. Then, a fault classification problem with real senor measurements collected from a power generation plant is evaluated.The results obtained are analyzed and explained, and implications of the proposed rule pruning strategy to the fuzzy classification system are discussed.

Rule learning and extraction using a hybrid neural network : a case study on fault detection and diagnosis

A hybrid network, based on the integration of Fuzzy ARTMAP (FAM) and the Rectangular Basis Function Network (RecBFN), is proposed for rule learning and extraction problems. The underlying idea for such integration is that FAM operates as a classifier to cluster data samples based on similarity, while the RecBFN acts as a “compressor” to extract and refine knowledge learned by the trained FAM network. The hybrid network is capable of classifying data samples incrementally as well as of acquiring rules directly from data samples for explaining its predictions. To evaluate the effectiveness of the hybrid network, it is applied to a fault detection and diagnosis task by using a set of real sensor data collected from a Circulating Water (CW) system in a power generation plant. The rules extracted from the network are analyzed and discussed, and are found to be in agreement with experts’ opinions used in maintaining the CW system.

Automatic rule induction in Arabic to English machine translation framework

This chapter addresses the exploitation of a supervised machine learning technique to automatically induce Arabic-to-English transfer rules from chunks of parallel aligned linguistic resources. The induced structural transfer rules encode the linguistic translation knowledge for converting an Arabic syntactic structure into a target English syntactic structure. These rules are going to be an integral part of an Arabic-English transfer-based machine translation. Nevertheless, a novel morphological rule induction method is employed for learning Arabic morphological rules that are applied in our Arabic morphological analyzer. To demonstrate the capability of the automated rule induction technique, we conducted rule-based translation experiments that use induced rules from a relatively small data set. The translation quality of the hybrid translation experiments achieved good results in terms of WER.

The Role of TAFE as an Equity Pathway to Social Inclusion, Employment, and to University : a submission from The Deakin University Social Work/Gordon TAFE Community Services Work, Geelong Based Project Team

Executive Summary

The Deakin University Social Work/Gordon TAFE Community Services Work Geelong Based Project Team (the Project Team) was assisted by Higher Education Partnership and Participation funding made available through Deakin University Participation and Partnerships Program (DUPPP) to carry out research and project work in 2012/13.

In the following submission to the House of Representatives Inquiry into the role of Technical and Further Education (TAFE) system and its operation, this Project Team seeks to establish a case for:

1. Funding to enable TAFE to continue as:

a) an equity pathway to social inclusion, employment, and to university, particularly in regional areas.
b) an integral complement to the University education sector to deliver on the ambitious objectives of the Federal Government’s widening participation agenda, as a mechanism to deliver the skills, knowledge and workforce needed now, and in the future, in the Australian economy.
2. Increased resources for separate and joint sector development
a) Publicly funded TAFEs need funding to be restored and increased to enable them to maintain the high quality education they provide and to maintain their successful work in supporting communities, regions and disadvantaged individuals to gain skills, training and employment.
b) Universities need increased funding to increase staffing levels and therefore free up teaching staff to spend the necessary time to develop relationships with and provide support to students. This is important for the achieving the goals of the widening participation agenda of increasing access without increasing attrition at the same time.
c) TAFEs and Universities need funding to do the work required to further develop and formalise diploma-degree pathways so that disadvantaged individuals can exit into employment at the diploma level or be supported in an efficient and seamless way to undertake further study.
3. Active use of localised and nuanced partnership approaches by education institutions. This includes:
• Cross teaching by TAFEs and Universities in courses that can be articulated, such as professional practice diplomas and degrees
• Programs negotiated and designed according to the needs of students in each location. TAFEs and Universities need resources in order to do this work
• Focus on regional centres where there is a particular opportunity for government to make an impact on TAFE pathways to employment and/or further education
• Workforce development in regional areas due to new industries is a particular area of need
4. Recognise and capitalise on the complementary and symbiotic nature of each sector’s skills, strengths and capacities.
The submission responds to the second, third and fifth points of the Terms of Reference of the Inquiry and is based on the research work carried out by the Project Team in 2012/13.

We provide evidence of Gordon TAFE in Geelong working as an equity mechanism in the particular case of the welfare/ community services diploma to social work degree pathway. The project team considers that there is a strong case for additional resourcing of TAFE to enable it to continue what it does well. TAFE is the key training and education sectorthe ‘education and social hub’that can successfully attract, retain, and graduate people who may not otherwise access education due to one or more combinations of:

1. having a low SES current or past background;
2. living in regional areas;
3. receiving interrupted primary and secondary education;
4. having disabilities;
5. being sole parents;
6. being from refugee backgrounds;
7. having English as an additional language/culture;
8. retrenchment from employment in dying industries;
9. short, medium and long term unemployment;
10. past and/or current caring roles;
11. marriage/relationship breakdowns;
12. domestic violence;
13. gender, class, age, race/ethnicity and dis/ability discriminations; and
14. socialised expectations and fears.

The recommendations in this submission are based on research findings about important similarities and differences between Gordon TAFE welfare and Deakin University social work students in Geelong, and their respective institutional organisations and contexts. The two institutions employ a repertoire of diverse administrative, teaching, learning and support approaches to meet different mission goals, requirements and needs.

A new framework with similarity reasoning and monotone fuzzy rule relabeling for fuzzy inference systems

A complete and monotonically-ordered fuzzy rule base is necessary to maintain the monotonicity property of a Fuzzy Inference System (FIS). In this paper, a new monotone fuzzy rule relabeling technique to relabel a non-monotone fuzzy rule base provided by domain experts is proposed. Even though the Genetic Algorithm (GA)-based monotone fuzzy rule relabeling technique has been investigated in our previous work [7], the optimality of the approach could not be guaranteed. The new fuzzy rule relabeling technique adopts a simple brute force search, and it can produce an optimal result. We also formulate a new two-stage framework that encompasses a GA-based rule selection scheme, the optimization based-Similarity Reasoning (SR) scheme, and the proposed monotone fuzzy rule relabeling technique for preserving the monotonicity property of the FIS model. Applicability of the two-stage framework to a real world problem, i.e., failure mode and effect analysis, is further demonstrated. The results clearly demonstrate the usefulness of the proposed framework.

Dorothy Gordon

Perhaps due to her later fame as the newspaper columnist and controversial radio personality known as “Andrea,” the early film career of Dorothy Gordon has been largely overlooked. Certainly her significant celebrity in Australia in the 1 960s overshadowed her comparatively lesser achievements as a silent cinema extra, actor, screenwriter, and art director. The recognition of her contribution to the Australian film industry is further diminished by the loss of her major work as star of Raymond Longford’s Hills of Hate (1926). In addition, her reputation as an entertaining raconteur fond of telling tall stories, especially about herself, leav es much room for doubt about the recorded detail of her early career in both Holly wood and Australian silent films.

Preparation and characterization of natural rubber/silica nanocomposites using rule of similarity in latex

Rule of similarity and latex compounding techniques were combined for the first time to prepare natural rubber/nanosilica (NR/SiO2) nanocomposite with core-shell nanosilica-poly (methyl methacrylate) (SiO 2-PMMA) particles and PMMA-modified natural rubber matrix (NR-PMMA). The microstructure of SiO2 and nanocomposites with different SiO 2 contents was characterized by fourier transform infrared spectroscopy (FTIR); the morphology of nanocomposites was investigated with scanning electron microscopy (SEM); the tensile strength was characterized by tensile testing machine and the thermal stability of composites was studied by thermal gravimetric analysis. Results showed that PMMA chains have successfully grafted onto the surface of SiO2, and the core-shell SiO 2-PMMA nanoparticles and NR-PMMA latex have been perfectly incorporated. SiO2-PMMA nanoparticles are evenly distributed over the NR matrix with an average size in the range of 60-100 nm at the low content (SiO2? 3 wt%), while aggregations are apparently observed when 5 wt% SiO2 is loaded. In addition, NR/SiO2 composities possess a considerable improvement in ageing resistance compared with the pure NR. The tensile strength of composite increases from 6.99 to 12.72 MPa, reaching the highest value at a 0.5 wt% SiO2 loading, and then the figure decreases gradually because of the aggregation of SiO2 nanoparticles. It is anticipated that the reported process is to provide a simple and economic way for preparing NR composites.

A possible anatomical and biomechanical explanation of the 10% rule used in the clinical assessment of prehensile hand movements and handed dominance

Abstract
A current doctrine in the dynamometric approach to determine lateralization of hand function states that in 10% of cases, the non-dominant hand will be stronger than the dominant hand. In this study, a novel MRI based modelling approach was applied to the first dorsal introsseus muscle (FDI), to determine whether the 10% rule may be applied to the FDI and may be partially explained by the arrangement of the anatomical components of the FDI.

Methods
Initially the force generated by the thumb segment during an isometric pushing task in the horizontal plane was measured from 25 strongly right-handed young males. Nine of these participants then had structural magnetic resonance imaging (sMRI) of the thumb and index osseous compartment. A modelling technique was developed to extract the muscle data and quantify the muscle line of action onto to the first metacarpal bone segment in order to quantify the muscle force at the point of momentary rotation – equilibrium.

Results
Eight of 25 subjects exhibited stronger force from the left hand. Six out of nine subjects from the MRI possessed significantly greater angles of attachment of the index osseous compartment on the left (non-dominant) hand. These six subjects also generated greater maximal isometric forces from the FDI of the left side. There was a significantly greater muscle volume for the right FDI muscle as compared to the left as measured from the reconstructed MRI slice data.

Conclusions
The calculated force produced by the muscle is related to the angle of attachment of the muscle to bone in the index osseous compartment. The MRI findings indicate that the 10% rule may be anatomically and biomechanically explained.