Output Categories Research Articles

An Introduction to Geographical Information Systems

Beginning with the discussion of what is GIS, this article explains the concept of geographical data and spatial data models. Two software systems, viz computer-aided design and image processing have been identified as the close relatives of GIS and the similarities and differences between these systems and GIS are described. The operational functions of GIS are described under the categories of data input, analysis, spatial query and data output. Various application areas of GIS alongwith specialized applications of GIS are discussed. Some of the GIS software packages have been mentioned. A brief view has been taken of the state of GIS in India and the article is concluded.

Estimation of a multiproduct cost function for physically frail older people.

Using the theory of multiproduct cost functions, this paper derives a cost function for physically frail older people living in private households, based on data collected between 1991 and 1995, for 472 subjects aged 65 years of age or over in four areas of England. The main characteristic of the cost function is that output categories are classified in terms of movements between different health states. These were measured by changes in activities of daily living (ADL) over 2 years, with 'low' ADL representing better functional ability than 'high' ADL. Empirical application of the approach, using four states defined in terms of worsening progression (stable low ADL; deteriorated or improved ADL; stable high ADL; deceased), indicated more favourable states were associated with lower costs. Multivariate analysis showed that the derived states were significantly related to costs which, when combined with variables indicating presence of particular chronic health conditions (diabetes or previous stroke), admission to continuing care accommodation and household structure, explained one-fifth of the variation in log average costs per week. Variables such as age, sex, carer input, social networks and level of cognitive functioning had no independent impact on costs. These findings could be used as a starting point for those interested in predicting the cost implications associated with the ageing population.

Generalization and exclusive allocation of credit in unsupervised category learning

A new way of measuring generalization in unsupervised learning is presented. The measure is based on an exclusive allocation, or credit assignment, criterion. In a classifier that satisfies the criterion, input patterns are parsed so that the credit for each input feature is assigned exclusively to one of multiple, possibly overlapping, output categories. Such a classifier achieves context-sensitive, global representations of pattern data. Two additional constraints, sequence masking and uncertainty multiplexing, are described; these can be used to refine the measure of generalization. The generalization performance of EXIN networks, winner-take-all competitive learning networks, linear decorrelator networks, and Nigrin's SONNET-2 network are compared.

Generalization and exclusive allocation of credit in unsupervised category learning

A new way of measuring generalization in unsupervised learning is presented. The measure is based on an exclusive allocation, or credit assignment, criterion. In a classifier that satisfies the criterion, input patterns are parsed so that the credit for each input feature is assigned exclusively to one of multiple, possibly overlapping, output categories. Such a classifier achieves context-sensitive, global representations of pattern data. Two additional constraints, sequence masking and uncertainty multiplexing, are described; these can be used to refine the measure of generalization. The generalization performance of EXIN networks, winner-take-all competitive learning networks, linear decorrelator networks, and Nigrin's SONNET-2 network are compared.

Cap reform: modelling supply response subject to the land set-aside

Abstract This paper uses duality theory to develop a model of European Community agriculture. The model is used to investigate the impact of the land set-aside provision of the recent package of reforms of the Common Agricultural Policy. We assume that producers chose output and variable input levels that maximize difference between revenue and variable cost. By including first-order conditions for the allocation of land across its uses, we impose that the observed allocations are profit-maximizing allocations. To overcome the problem of incorporating many outputs into an estimable production structure, we imposed a priori the restriction that the technology was weakly separable in major categories of outputs. With this restriction, it was possible to model production decisions in stages using consistent aggregates in the latter stages.

Cap reform: modelling supply response subject to the land set‐aside

AbstractThis paper uses duality theory to develop a model of European Community agriculture. The model is used to investigate the impact of the land set‐aside provision of the recent package of reforms of the Common Agricultural Policy. We assume that producers chose output and variable input levels that maximize difference between revenue and variable cost. By including first‐order conditions for the allocation of land across its uses, we impose that the observed allocations are profit‐maximizing allocations. To overcome the problem of incorporating many outputs into an estimable production structure, we imposed a priori the restriction that the technology was weakly separable in major categories of outputs. With this restriction, it was possible to model production decisions in stages using consistent aggregates in the latter stages.

Probabilistic neural network prediction of ascites in broilers based on minimally invasive physiological factors

A Probabilistic Neural Network (PNN) was trained to predict ascites in broilers based on minimally invasive inputs (i.e., physiological factors that do not require the death of the bird). A PNN is a supervised, three-layer, artificial neural network that classifies input patterns (e.g., physiological data) into specific output categories (e.g., ascites or no ascites). The PNN inputs were O2 level in the blood, body weight, electrocardiogram (ECG), hematocrit, S wave, and heart rate of individual birds. These data were from three experiments that have been described previously (Roush et al., 1996a,b). The three data sets were pooled into a combined data set for a total of 170 observations. From the pooled data, a training set (117 birds), a calibration set (17 birds), and a verification set (36 birds) were extracted. The PNN was trained on the training data set. To prevent the PNN from overfitting the training data, the neural network was evaluated on its ability to make correct predictions of the calibration data set. At the point at which the neural network made the highest number of correct classifications for the calibration data set, the trained neural network was saved on the computer. When the PNN was applied to the complete data set, the sensitivity or proportion of the birds with ascites that the PNN correctly diagnosed was 0.97 (75/77 birds). The specificity or proportion of birds that the PNN made a correct diagnosis of not having ascites was 0.98 (91/93 birds). When the PNN was applied to the verification data set, which was not subjected to neural network training, the sensitivity was 0.95 (19/20) and the specificity was 0.88 (14/16 birds). Use of models developed with artificial neural networks may enhance the diagnosis of ascites in broilers. The results may be useful in choosing and developing broiler strains that do not have a propensity for ascites.

EFFECTS OF DATA PREPROCESSING ON THE PERFORMANCE OF ARTIFICIAL NEURAL NETWORKS FOR DAIRY YIELD PREDICTION AND COW CULLING CLASSIFICATION

The effect of data preprocessing on the learning ability of artificial neural networks was investigated with regardto the impact of distributing the input vectors uniformly with respect to the output categories in the training data set. Theanalyses were performed for neural networks dedicated to (1) dairy cow culling classification and (2) milk yield prediction.The two types of neural network used for culling classification were backpropagation and learning vector quantization. Foryield prediction, backpropagation was used. The study was repeated with several architectures for both types of network.Preprocessing of data did not have a large impact on the general performance of the networks, but did affect the results foreach output category. The effects were more pronounced in the categories containing less frequent events, for which theresults always improved. For the categories with larger number of records, balancing the data degraded the results. Therespective improvements and degradation of the results occurred for both prediction and classification, with the two types ofneural networks, and with all architectures tested. However, the magnitude of the effects varied with the type of neuralnetwork and with the architecture. The results of this study indicate that, in general, the distribution of outputs influences thelearning process of neural networks for both types of application. The results also suggest that the types of outputdistribution required for the training of neural nets may depend on the specifics of each problem.

THE DEMAND FOR DOMESTIC OUTPUT AND IMPORTS IN THE U.K.: A DYNAMIC SYSTEMS APPROACH

This paper is concerned with the allocation of total final expenditure between different categories of domestic net output and imported goods and services. In contrast to previous studies, the authors show that the use of net output data means that allowance has to be made for both consumer preference and the structure of production technologies, since the latter determine the demand for intermediate products. They argue that the full complexity of the relationships between different products can only be properly captured through the simultaneous estimation of a disaggregated sectoral model and report econometric estimates using U.K. quarterly data over a period from 1968 to 1991. Copyright 1996 by Blackwell Publishers Ltd and The Victoria University of Manchester

On the match tracking anomaly of the ARTMAP neural network

This article analyses the match tracking anomaly (MTA) of the ARTMAP neural network. The anomaly arises when an input pattern exactly matches its category prototype that the network has previously learned, and the network generates a prediction (through a previously learned associative link) that contradicts the output category that was selected upon presentation of the corresponding target output. Carpenter et al. claimed that such an anomalous situation will never arise if the (binary) input vectors have the same number of 1s (Carpenter et al., 1991, Neural Networks, 4, 565–588). This paper shows that such situations can in fact occur. The timing according to which inputs are presented to the network in each learning trial is crucial: if the target output is presented to the network before the corresponding input pattern, certain pattern sequences will lead the network to the MTA. Two kinds of MTA are distinguised: one that is independent of the choice parameter (β) of the ART b module, and another that is not. Results of experiments that were carried out on a machine learning database demonstrate the existence of the match tracking anomaly as well as support the analytical results presented here.

EXPERIENCE WITH THE IBM ZISC036 NEURAL NETWORK CHIP

The new IBM Zero Instruction Set Computer (ZISC) provides a radial basis function neural network. The first generation chip (ZISC036) allows for 64 8-bit inputs, 36 RBF neurons in the middle layer, and up to 16383 possible output categories. Forward processing takes 4μs with a 20MHz clock. Cascading multiple chips increases the number of available RBF’s with no increase in processing time. The chip also executes a learning algorithm. We report on tests of the ZISC with a high energy physics related task.

A neural network for unsupervised categorization of multivalued input patterns: an application to satellite imaee clustering

This article presents an implementation of an artificial neural network (ANN) which performs unsupervised detection of recognition categories from arbitrary sequences of multivalued input patterns. The proposed ANN is called Simplified Adaptive Resonance Theory Neural Network (SARTNN). First, an Improved Adaptive Resonance Theory 1 (IARTl)-based neural network for binary pattern analysis is discussed and a Simplified ARTl (SART1) model is proposed. Second, the SARTl model is extended to multivalued input pattern clustering and SART” is presented. A normalized coefficient which measures the degree of match between two multivalued vectors, the Vector Degree of Match (VDM), provides SARTNN with the metric needed to perform clustering. Every ART architecture guarantees both plasticity and stability to the unsupervised learning stage. The SARTNN plasticity requirement is satisfied by implementing its attentional subsystem as a self-organized, feed-forward, flat Kohonen’s ANN (KANN). The SARTNN stability requirement is properly driven by its orienting subsystem. SARTNN processes multivalued input vectors while featuring a simplified architectural and mathematical model with respect to both the ARTl and the AkT2 models, the latter being the ART model fitted to multivalued input pattern categorization. While the ART2 model exploits ten user-defined parameters, SARTNN requires only two user-defined parameters to be run: the first parameter is the vigilance threshold, p, that affects the network’s sensibility in detecting new output categories, whereas the second parameter, T, is related to the network's learning rate. Both parameters have an intuitive physical meaning and allow the user to choose easily the proper discriminating power of the category extraction algorithm. The SARTNN performance is tested as a satellite image clustering algorithm. A chromatic component extractor is recommended in a satellite image preprocessing stage, in order to pursue SARTNN invariant pattern recognition. In comparison with classical clustering algorithms like ISODATA, the implemented system gives good results in terms of ease of use, parameter robustness and computation time. SARTNN should improve the performance of a Constraint Satisfaction Neural Network (CSNN) for image segmentation. SARTNN, exploited as a self-organizing first layer, should also improve the performance of both the Counter Propagation Neural Network (CPNN) and the Reduced connectivity Coulomb Energy Neural Network (RCENN).

Telecommunications technology, sectoral prices and international competitiveness

The impact of technological advances in telecommunication service production over the interval 1963–1982 upon the prices of telecommunication services and other categories of commodity output is calculated and compiled to show effects upon broader price indices, the wholesale price index (WPI) and the consumer price index (CPI). The analysis reveals the impact of telecommunications advances on US sectoral prices to be significant. Moreover, technical advances in US telecommunications have played an important role in international competitiveness and trade. Because existing technological advances have not become fully diffused in the USA these results raise questions concerning the diffusion process, and suggest there are social benefits to accelerated diffusion.

PROGRAM BUDGETING: COSTS AND BENEFITS

: This article focuses on the divergence between the principle and the reality of program budgeting. The principle is that program budgeting is based on the “output” classification of expenditure. In reality, however, program budgets are more like hybrids between the organisational and functional classifications of expenditure. The reasons for this divergence between principle and reality are analysed by examining the costs and benefits of the output classification of expenditure, contrasted to organisational/functional classification, for the key purposes of budgeting—namely control, estimation and expenditure review and appraisal. It is concluded that, although the output classification of expenditure offers significant benefits from the standpoint of expenditure review and appraisal, it is inferior to the organisational/functional classification from a cost benefit point of view. This is essentially because of the very substantial costs of obtaining reliable information on actual expenditure in terms of output categories, relative to the lower costs of obtaining organisational/functional information. With respect to the other purposes of budgeting, two conclusions are drawn. First, output categories are much less suited to the estimation of forward expenditure than are organisational/functional categories. Second, the use of output categories for control (i.e. expenditure instruction) purposes is inappropriate, principally because it is unnecessary and creates unwarranted expenditure inflexibilities. It is argued, however, that this deficiency of the output categories for control purposes is a reflection of the general inappropriateness of prescriptive program budgets (whether based on output or organisationa/functional categories).

The calculation of operational performance ratings

We defined a production unit (PU) as a purposeful system that converts resources (inputs) into goods and/or services (outputs). In the manufacturing and service sectors of an economy, there are many different types of PUs ranging in size from individuals to large organizations. A PU's overall competitiveness depend is very much on its operational performance, that is its efficiency with which it produces goods and services. For operational performance control, the development of a reliable measurement system is essential. The objective of the present paper is to develop a procedure to calculate relative operational performance ratings of a set of PUs. The criteria we set for the ratings include intuitive appeal, computational ease, and ability to reflect different managerial perspectives and priorities. We demonstrate the application of the procedure using a data set from the recent literature, which involves twelve input and two output categories in a manufacturing setting.

Federal funding and R&D output in US industrial laboratories

The factors associated with inter-R&D laboratory differences in output from the R&D processes are investigated with emphasis on the influence of federal R&D on the composition of this output. An attempt is made to describe the various categories of innovative output characterizing industrial R&D and to investigate factors affecting the distribution of these activities among industrial R&D laboratories. It is shown that federal R&D allocations significantly redirect a laboratory's output, skewing it toward activities with public-good characteristics. >

Performance measurement in universities

Universities have traditionally possessed as their primary objectives the pursuit and the transmission of knowledge. During the 1980's, financial pressures on British universities have greatly increased, all budgets being cut in real terms, dramatically so for a handful of universities. The current emphasis on effectiveness and efficiency has naturally led on to a thorough examination of measures of performance. What is the organisational effectiveness of a University (and of a constituent department) and how do we measure it? Universities may be regarded as offering three major categories of output: (i) highly qualified manpower; (ii) research and scholarship; (iii) various other social benefits e.g. contribution to ‘national culture’ or valuable contributions to the life of their local communities. Clearly some of these outputs are much easier to measure than others. The paper will examine the methods used to measure output and to relate categories of output to the relevant inputs, including the concept of ‘the social rate of return’. Performance indicators will be included under three major categories: • (i) Internal, e.g. first degree graduation rates, success rates of higher degrees, attraction of research funds. • (ii) External, e.g. acceptability of graduates in employment, staff publications, patents. • (iii) Operating, e.g. unit costs, staff/student ratios, staff workloads. The third category includes a number of traditional problem areas for OR/MS work such as establishing service levels for libraries and computer systems. The overall approach clearly possesses considerable implications for the whole philosophy and culture of universities and for individual members of staff. The paper will touch on these aspects too.

Incremental learning of large phonetic neural networks from smaller subnets

Time delay neural networks (TDNNs) [Waibel et al., ATR‐TR‐0006 (1987)] have recently been shown to achieve excellent recognition performance for difficult phonetic discrimination tasks (e.g., voiced stops). This was achieved in part by the TDNNs' ability to not only activate the correct output category, but also to inhibit all incorrect outputs. The disadvantage of this property is that training larger networks with many more output categories (e.g., all consonants) becomes nontrivial, as all categories have to be incorporated in the learning process requiring excessive amounts of training. Several techniques are presented that overcome this problem by exploiting the hidden structure of previously learned smaller neural nets to train larger nets incrementally in comparatively short training runs. Experimental results show that the resulting larger networks aimed at stop consonants, at voiced stops and nasals, and at all consonants achieve recognition scores (95%–99%) as high as the smaller subnetworks from which they were constructed.

Research Productivity and Concentration Among Agricultural Economics Faculty in the Northeast

Research activities of academic agricultural economists in the Northeast Region are examined. Selected categories of research output are presented. Interregional comparisons in research productivity are made between: 1) faculty employed in the Northeast and those employed elsewhere and 2) faculty educated in the Northeast and those educated elsewhere. Intraregional comparisons of faculty employed in the Northeast by region of education are also presented. Regional differences in research resources and concentration are examined and offered as factors contributing to research productivity differences. Results indicate few significant long-run differences in research productivity and concentration in the Northeast relative to that in other regions.

The World Politics of World Hunger: A Study in Biopolitical Interactions

Human nutrition provides a subject area in which biology and politics interact in two respects. First, nutrition is a factor in human political behavior that may affect both individual decision-making performance and mass participation. Specific conditions, such as hypoglycemia, may affect individual performance, while inadequate energy intakes and qualitative deficiencies may influence mass behavior. Second, public policy affects the nutritional status of individuals and populations. Agricultural policy, food trade policy, nutritional intervention programs, foreign and domestic food aid, and social welfare policies are among the categories of governmental outputs that help determine nutritional conditions in populations and their component groups. Nutritional status, in turn, has feedback effects through mass behavior on political inputs. Thus, human nutrition presents a problem of biopolitical interactions that involves a mutually interdependent relationship with cybernetic properties.