Bayesian networks for enhancement of requirements engineering: a literature review

Requirements analysis is the software engineering stage that is closest to the users’ world. It also involves tasks that are knowledge intensive. Thus, the use of Bayesian networks (BNs) to model this knowledge would be a valuable aid. These probabilistic models could manage the imprecision and ambiguities usually present in requirements engineering (RE). In this work, we conduct a literature review focusing on where and how BNs are applied on subareas of RE in order to identify which gaps remain uncovered and which methods might engineers employ to incorporate this intelligent technique into their own requirements processes. The scarcity of identified studies (there are only 20) suggests that not all RE areas have been properly investigated in the literature. The evidence available for adopting BNs into RE is sufficiently mature yet the methods applied are not easily translatable to other topics. Nonetheless, there are enough studies supporting the applicability of synergistic cooperation between RE and BNs. This work provides a background for understanding the current state of research encompassing RE and BNs. Functional, non-functional and -ilities requirements artifacts are enhanced by the use of BNs. These models were obtained by interacting with experts or by learning from databases. The most common criticism from the point of view of BN experts is that the models lack validation, whereas requirements engineers point to the lack of a clear application method for BNs and the lack of tools for incorporating them as built-in help functions.     

LPS: A Language Prototyping System Using Modular Monadic Semantics

This paper describes LPS, a Language Prototyping System that facilitates the modular development of interpreters from semantic building blocks. The system is based on the integration of ideas from Modular Monadic Semantics and Generic Programming.To define a new programming language, the abstract syntax is described as the fixpoint of non-recursive pattern functors. For each functor an algebra is defined whose carrier is the computational monad obtained from the application of several monad transformers to a base monad. The interpreter is automatically generated by a catamorphism or, in some special cases, a monadic catamorphism.The system has been implemented as a domain-specific language embedded in Haskell and we have also implemented an interactive framework for language testing.     

Robot arm reaching through neural inversions and reinforcement learning

We present a neural method that computes the inverse kinematics of any kind of robot manipulators, both redundant and non-redundant. Inverse kinematics solutions are obtained through the inversion of a neural network that has been previously trained to approximate the manipulator forward kinematics. The inversion provides difference vectors in the joint space from difference vectors in the workspace. Our differential inverse kinematics (DIV) approach can be viewed as a neural network implementation of the Jacobian transpose method for arm kinematic control that does not require previous knowledge of the arm forward kinematics. Redundancy can be exploited to obtain a special inverse kinematic solution that meets a particular constraint (e.g. joint limit avoidance) by inverting an additional neural network The usefulness of our DIV approach is further illustrated with sensor-based multilink manipulators that learn collision-free reaching motions in unknown environments. For this task, the neural controller has two modules: a reinforcement-based action generator (AG) and a DIV module that computes goal vectors in the joint space. The actions given by the AG are interpreted with regard to those goal vectors.     

Efficient Hardware/Software Implementation of an Adaptive Neuro-Fuzzy System

This paper describes the development of efficient hardware/software (HW/SW) neuro-fuzzy systems. The model used in this work consists of an adaptive neuro-fuzzy inference system modified for efficient HW/SW implementation. The design of two different on-chip approaches are presented: a high-performance parallel architecture for offline training and a pipelined architecture suitable for online parameter adaptation. Details of important aspects concerning the design of HW/SW solutions are given. The proposed architectures have been implemented using a system-on-a-programmable-chip. The device contains an embedded-processor core and a large field programmable gate array (FPGA). The processor provides flexibility and high precision to implement the learning algorithms, while the FPGA allows the development of high-speed inference architectures for real-time embedded applications.     

Temporal aggregation, systematic sampling, and the Hodrick-Prescott filter

The time aggregation properties of the Hodrick-Prescott (HP) filter, which decomposes a time series into trend and cycle, are analyzed for the case of annual, quarterly, and monthly data. Aggregation of the disaggregate components cannot be obtained as the exact result from direct application of an HP filter to the aggregate series. Employing several criteria, HP decompositions for different levels of aggregation that provide similar results can be found. The aggregation is guided by the principle that the period associated with the frequency for which the filter gain is should not be altered. This criterion is intuitive and easy to apply. It is shown that it is approximated, to the first order, by an already proposed empirical rule and that alternative, more complex criteria yield similar results. Moreover, the values of the smoothing parameter of the HP filter that provide results which are approximately consistent under aggregation are considerably robust with respect to the ARIMA model of the series. Aggregation is found to perform better for the case of temporal aggregation than for systematic sampling. The desirability of exact aggregation consistency is investigated. A clarification concerning the supposed spuriousness of the cycles obtained by the HP filter is discussed.     

In situ autolysis of mouse brain: ultrastructure of mitochondria and the function of oxidative phosphorylation and mitochondrial DNA

The effect of in situ autolysis on cerebral mitochondrial structure and function has been investigated. Mice (n = 9) were sacrificed and stored for up to 24 h under unfavorable post-mortem conditions at 25 degrees C. At different time intervals groups of three animals were submitted to post-mortem dissection and tissue from different regions of the brain was used for the preparation of "free" and synaptosomal mitochondria. On electron microscopic examination, the post-mortem period had no significant influence on mitochondrial morphology and enzymatic activities of complexes I-V of the mitochondrial oxidative phosphorylation system were still present in all the mitochondrial preparations from different regions of the brain, albeit at a reduced levels. Degradation of mitochondrial DNA was virtually absent from mitochondrial preparations during the 24-h period of autolysis, as shown by the presence of intact DNA by Southern blot and PCR analysis. Based on these results, alterations in mitochondrial DNA and deficiencies of mitochondrial respiratory complexes I-V can be recognized in cerebral tissue even after 24 h of unfavorable post-mortem storage conditions.     

A consensus model for group decision making problems with linguistic interval fuzzy preference relations

Sometimes, we find decision situations in which it is difficult to express some preferences by means of concrete preference degrees. In this paper, we present a consensus model for group decision making problems in which the experts use linguistic interval fuzzy preference relations to represent their preferences. This model is based on two consensus criteria, a consensus measure and a proximity measure, and on the concept of coincidence among preferences. We compute both consensus criteria in the three representation levels of a preference relation and design an automatic feedback mechanism to guide experts in the consensus reaching process.     

Tool wear monitoring using neuro-fuzzy techniques: a comparative study in a turning process

Tool wear detection is a key issue for tool condition monitoring. The maximization of useful tool life is frequently related with the optimization of machining processes. This paper presents two model-based approaches for tool wear monitoring on the basis of neuro-fuzzy techniques. The use of a neuro-fuzzy hybridization to design a tool wear monitoring system is aiming at exploiting the synergy of neural networks and fuzzy logic, by combining human reasoning with learning and connectionist structure. The turning process that is a well-known machining process is selected for this case study. A four-input (i.e., time, cutting forces, vibrations and acoustic emissions signals) single-output (tool wear rate) model is designed and implemented on the basis of three neuro-fuzzy approaches (inductive, transductive and evolving neuro-fuzzy systems). The tool wear model is then used for monitoring the turning process. The comparative study demonstrates that the transductive neuro-fuzzy model provides better error-based performance indices for detecting tool wear than the inductive neuro-fuzzy model and than the evolving neuro-fuzzy model.     

Analysis of the cyclic short-term variation of indoor power line channels

The purpose of this paper is to analyze the properties of indoor power line communications channels when they are used for broadband transmission. It will be shown that these channels exhibit a short-term variation, due to the fact that high-frequency parameters of electrical devices depend on the instantaneous amplitude of mains voltage. This phenomenon leads to a channel model proposal based on a linear periodically time-varying system and cyclostationary random noise. A measurement system, designed to capture these time variations in real power networks, and some of the most significant results obtained with it, are presented. In addition, a statistical analysis of the measurements performed in several indoor power line scenarios is included, so that the relevance of time variations in actual channels can be evaluated.     

Super‐resolution of the undersampled and subpixel shifted image sequence by a neural network

Numerous approaches to super‐resolution (SR) of sequentially observed images (image sequence) of low resolution (LR) have been presented in the past two decades. However, neural network methods are almost ignored for solving SR problems. This is because the SR problem traditionally has been regarded as the optimization of an ill‐posed large set of linear equations. A designed neural network based on this has a large number of neurons, thereby requiring a long learning time. Also, the deduced cost function is overly complex. These defects limit applications of a neural network to an SR problem. We think that the underlying meaning of the SR problem should refer to super‐resolving an imaging system by image sequence observation, instead of merely improving the image sequence itself. SR can be regarded as a pattern mapping from LR to SR images. The parameters of the pattern mapping can be learned from the imaging process of the image sequence. This article presents a neural network for SR based on learning from the imaging process of the image sequence. In order to speed up the convergence, we employ vector mapping to train the neural network. A mapping vector is composed of some neighbor subpixels. Such a well‐trained neural network has powerful generalization ability so that it can be used directly to estimate the SR image of the other image sequences without learning again. Our simulations show the effectiveness of the proposed neural network. © 2004 Wiley Periodicals, Inc. Int J Imaging Syst Technol 14, 8–15, 2004; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.20001