Abstract
Good documentation is crucial for the success of frameworks. In this research, a new documenting approach is proposed combining existing document styles in a kind of "minimalist" framework manual with a special emphasis on framework understandability and usability, rather than on describing framework design. Benefits and drawbacks are evaluated from frameworks of different domains and complexity. © 2000 ACM.

Abstract
This paper describes a Load Allocation model to be used in a DMS environment. A process of rough allocation is initiated, based on information on actual measurements and on data about installed capacity and power and energy consumption at LV substations. This process generates a fuzzy load allocation, which is then corrected by a fuzzy state estimator procedure in order to generate a crisp power flow compatible set of load allocations, coherent with available real time measurements recorded in the SCADA.

Abstract
The Cellular Neural Networks (CNN) model is now a paradigm of cellular analogue programmable multidimensional processor array with distributed local logic and memory. CNNs consist of many parallel analogue processors computing in real time. One desirable feature is that these processors arranged in a two dimensional grid only have local connections, which lend themselves easily to VLSI implementations. In this paper, we present a new algorithm for motion estimation using CNN. We start from a mathematical viewpoint (i.e., statistical regularisation based on Markov Random Field, (MRF)) and proceed by mapping the algorithm onto a cellular neural network. Because of the temporal dynamics inherent in the cells of the CNN it is well suited to processing time-varying images. A robust motion estimation algorithm is achieved by using a spatio-temporal neighbourhood for modelling pixel interactions.

Abstract
Controller Area Network (CAN) is a fieldbus network suitable for small-scale Distributed Computer Controlled Systems, being appropriate for transferring short real-time messages. Nevertheless, it must be understood that the continuity of service is not fully guaranteed since it may be disturbed by temporary periods of network inaccessibility [1]. In this paper, such temporary periods of network inaccessibility are integrated in the response time analysis of CAN networks. The achieved results emphasise that in the presence of temporary periods of network inaccessibility, a CAN network is not able to provide different integrity levels to the supported applications, since errors in low priority messages interfere with the response time of higher priority message streams.

Abstract
Naive Bayes is a well known and studied algorithm both in statistics and machine learning. Bayesian learning algorithms represent each concept with a single probabilistic summary. In this paper we present an iterative approach to naive Bayes. The iterative Bayes begins with the distribution tables built by the naive Bayes. Those tables are iteratively updated in order to improve the probability class distribution associated with each training example. Experimental evaluation of Iterative Bayes on 27 benchmark datasets shows consistent gains in accuracy. Moreover, the update schema can take costs into account turning the algorithm cost sensitive. Unlike stratification, it is applicable to any number of classes and to arbitrary cost matrices. An interesting side effect of our algorithm is that it shows to be robust to attribute dependencies.

Abstract
In a classification problem, improved accuracy can be obtained in many situations by using the combination of several classifiers instead of a single one. In [10], the error reduction that can be obtained by combining unbiased classifiers with independent errors using a simple average, was derived. We present an extension of this result by finding the improvement obtained when combining classifiers using weighted average. We also prove that for unbiased classifiers with independent errors the best combination of N classifiers corresponds to a weighted average, where the combination coefficient of each classifier is equal to 1/N. This means that in these cases the simple average should be used. We present experiments illustrating our results.