Partial Decisions Research Articles

BioPrediction-RPI: Democratizing the prediction of interaction between non-coding RNA and protein with end-to-end machine learning

Machine Learning (ML) algorithms have been important tools for the extraction of useful knowledge from biological sequences, particularly in healthcare, agriculture, and the environment. However, the categorical and unstructured nature of these sequences requiring usually additional feature engineering steps, before an ML algorithm can be efficiently applied. The addition of these steps to the ML algorithm creates a processing pipeline, known as end-to-end ML. Despite the excellent results obtained by applying end-to-end ML to biotechnology problems, the performance obtained depends on the expertise of the user in the components of the pipeline. In this work, we propose an end-to-end ML-based framework called BioPrediction-RPI, which can identify implicit interactions between sequences, such as pairs of non-coding RNA and proteins, without the need for specialized expertise in end-to-end ML. This framework applies feature engineering to represent each sequence by structural and topological features. These features are divided into feature groups and used to train partial models, whose partial decisions are combined into a final decision, which, provides insights to the user by giving an interpretability report. In our experiments, the developed framework was competitive when compared with various expert-created models. We assessed BioPrediction-RPI with 12 datasets when it presented equal or better performance than all tools in 40% to 100% of cases, depending on the experiment. Finally, BioPrediction-RPI can fine-tune models based on new data and perform at the same level as ML experts, democratizing end-to-end ML and increasing its access to those working in biological sciences.

Strategic fit in global production networks – A decision support model for strategic configuration of global production networks

The configuration of global production networks is a highly complex management task. It involves decisions like site selection, site specialization, and long-term investments. Entering new markets, securing resource access and realizing cost advantages are only examples for the vast variety of strategic motives overlapping in these decisions. However, current material bottlenecks, geopolitical risks, and restrictions due to COVID-19 are increasingly bringing risk mitigation into the focus of decision-makers. In practice, an optimal configuration of the network in line with the production strategy, which is referred to as strategic fit, is rarely considered systematically. Approaches from academia do provide a variety of decision support models. However, these either consider the production network very aggregated and qualitative and are thus limited in their explanatory power, or they only focus on partial decisions, which are only evaluated based on costs. The aim of this paper is, therefore, to provide a decision support model for the configuration of the production network in line with the production strategy. The model derives an optimal, company-specific configuration based on strategic capabilities that constitute both the production strategy and the internal and external company environment. The procedure comprises six steps and is divided into an as-is analysis and a target derivation. In the as-is analysis, the current configuration is first recorded with description models. Subsequently, current strategic capabilities as well as the company environment are captured. Based on this data, a fuzzy inference system assesses the actual strategic capabilities of individual plants and the entire network. The strategic capability gap forms the starting point for the target derivation. First, a strategically fitting network phenotype is selected to provide guidelines for further configuration. Then, capability-oriented design principles are used to further elaborate the phenotypes. Subsequently, the resulting alternatives are evaluated with a fuzzy-TOPSIS evaluation model.

Pengaruh Variasi Produk Dan Citra Merek Terhadap Keputusan Pembelian Ulang Pada Susu Setia Di Tenggarong

Abstract: 
 The purpose of this study was to analyze the effect of product variations and brand image simultaneously, to analyze product variations and brand image to partial repurchase decisions, and to determine the independent variables that had the most dominant influence on repurchase decisions at Susu Setia. The analytical tool used in this research is multiple regression analysis. Samples were taken using the Slovin formula as many as 99.61 or 100 people.
 The results showed that the variable product variation and brand image had a simultaneous and partial effect on repurchase decisions at Susu Setia in Tenggarong. In addition, the Brand Image variable has the most dominant influence as seen from the B coefficient value of the Brand Image which is the largest compared to other variables. Based on the characteristics of respondents in the categories of gender, occupation, and age, it is proven that consumers are dominated by the female gender, millennial age with student status. Thus, the promotions carried out by Susu Loyal to increase repurchase decisions are targeted at this segment.
 
 Keywords: Product Variation, Brand Image and Repurchase Decision

Komplexität von Kaufentscheidungen im touristischen Kontext: Erkenntnisse für andere Dienstleistungsdomänen

Tourist purchasing decisions are multi-layered and complex, due to the combination of a number of special features: (1) Decisions take place at a distance and are therefore risky; (2) Decisions are characterized by portfolio logic with a large number of interdependent and often substitutive partial decisions (often also depending on decisions in the social context, e.g. fellow travelers), (3) tourism production is subject to a household production logic with a high degree of own operations in a highly emotional service context due to the absence of routine. In this paper we briefly summarize the central findings on tourism purchasing decisions based on the current state of literature and present an integrative generic purchasing decision model as a response to the prevailing partial or very specific and not very generalizable overall models. From this we then derive conclusions for (tourism) service industries and other service domains.

Risk assessment in water resources planning under climate change at the Júcar River basin

Abstract. Climate change and its possible effects on water resources has become an increasingly near threat. Therefore, the study of these impacts in highly regulated systems and those suffering extreme events is essential to deal with them effectively. This study responds to the need for an effective method to integrate climate change projections into water planning and management analysis in order to guide the decision-making, taking into account drought risk assessments. Therefore, this document presents a general and adaptive methodology based on a modeling chain and correction processes, whose main outcomes are the impacts on future natural inflows, a drought risk indicator, and the simulation of future water storage in the water resources system (WRS). This method was applied in the Júcar River basin (JRB) due to its complexity and the multiannual drought events it suffers recurrently. The results showed a worrying decrease in future inflows, as well as a high probability (≈80 %) of being under 50 % of total capacity of the WRS in the near future. However, the uncertainty of the results was considerable from the mid-century onwards, indicating that the skill of climate projections needs to be improved in order to obtain more reliable results. Consequently, this paper also highlights the difficulties of developing this type of method, taking partial decisions to adapt them as far as possible to the basin in an attempt to obtain clearer conclusions on climate change impact assessments. Despite the high uncertainty, the results of the JRB call for action and the tool developed can be considered as a feasible and robust method to facilitate and support decision-making in complex basins for future water planning and management.

Displacement, Legal Mobilization, and Disasters: Trial Courts and Legal Process

The process of litigation is part of the decentered complex governance structure of risks and disaster. The process highlights problems that other institutions deflect. Courts are also part of defining problems in governing disaster. Scholarship drawing on legal decisions, including concerning environmental decision making, often relies upon the final decision from a highest court of appeal. Most cases settle, and courts make temporary decisions that other courts subsequently overturn or vacate. This process also names and deflects problems. Therefore, looking only to highest courts of appeals misstates how courts participate in governance. However, the process of litigating in lower courts is complex to follow, with partial decisions and settlements. This paper traces the histories of key housing assistance cases taken after Hurricane Katrina, following their multiple iterations as well as how later cases drew upon them, through 2018. Although initial decisions evinced concern for those who had lost their homes, flexible legal standards and limits on groups' ability to litigate allowed courts to limit government agencies’ accountability in court. This paper argues for integrating courts into the governance of risks and hazards, and for following trial courts and process.

Awards, Orders and Other Types of Decisions

Awards, Orders and Other Types of Decisions

Towards a Real Reconciliation of Centralised and Decentralised Approach in Legal Regulation: Application of an AI-Toolkit to a Problem of Jurisprudence

We present results from the implementation of an approach and a toolkit overcoming the limitations of the existing approaches by which law creates mechanisms for regulating socio-economic interests. In the centralized approach the rules come down from the rulers to the ruled ones. That limits the range of interested individuals to establish and implement regulatory mechanisms. In the decentralized approach, if necessary, rulers delegate certain individuals to create rules ad hoc, expanding the range of subjects of decision. This presupposes possibility of power abuse and take conflicting decisions in analogous cases.These disadvantages have been identified back in the ancient Rome. For their overcoming the systems of ius civile and ius honorarium were combined (II B.C.-I A.C.). Such attempts continue, for example within the EU law. A model that overcomes these disadvantages hasn`t been applied so far, because optimal and socially-efficient combination of centralized and decentralized approaches presupposes specific mechanisms. Their establishment and implementation should cover all interested individuals by coordinating and integrating the partial decisions, guaranteeing an optimal adaptability of the society towards the environment. Such mechanisms are achievable by the application of a cognitive approach, methodology and the G-space architecture for identifying inaccurately or mistakenly formulated motivational mechanisms in legislation. We describe how this set of cognitive instruments could be used to correct mistakes in existing legislation and design new, subject to ex-ante defined purposes. We show that the approach, methodology and G-space architecture extend the boundaries of the inter-disciplinary area of Artificial Intelligence in Law.

Multiple Antenna-Aided Cascaded Energy and Matched Filter Detector for Cognitive Radio Networks

A new multiple antenna-aided cascaded energy and matched filter detector (MCEM) for cognitive radio networks is proposed. Our proposed scheme is equipped with multiple receive antennas employing energy efficient energy detector (ED) and reliable matched filter (MF) to mitigate the channel fading and increase the detection performance of the secondary users (SUs). Partial decisions are first made by using cascaded energy and matched filter detector (CEM) at each antenna and then the final collaborative decision is made based on those partial decisions. The probability of detection/false alarm of the proposed scheme is presented in terms of the complementary receiver operating characteristics (ROCs). The performance of MCEM scheme is explored via MATLAB simulations that implement the clear channel assessment (CCA) modules for IEEE 802.15.4. Simulation results showed that our proposed MCEM scheme improves the detection performance and is more energy efficient as compared to CEM scheme employing single antenna and other conventional schemes.

Why Should Justice Be Seen to Be Done?

A well-known maxim instructs that justice should be seen to be done. When “seen” is understood in the sense of “observed”, the maxim is easily defended: open court proceedings protect against arbitrary and partial decisions. However, when “seen” is understood in the sense of “seem,” the maxim is more puzzling, since it is not obvious why courts should concern themselves with people's perceptions that justice has been done. This article addresses this issue, with a particular focus on the social and other benefits that result when judges observe procedures that are widely regarded as fair, especially in criminal trials. The article draws on empirical studies in social psychology that show that when legal authorities treat people in ways that accord with “lay” procedural expectations, they are more likely to view the authorities as legitimate, to cooperate with them, and to obey the law out of an internalized sense of obligation. The article explores the moral significance of these empirical findings, arguing that it would be superficial to see them as a recipe for social stability. The deeper truth conveyed by the empirical research is that relating to people in ways that are widely perceived to be fair is a way for authorities to engage people's moral sentiments and to enliven their virtuous capacity to put aside considerations of self-interest so as to do what is right. This dynamic provides a sound moral foundation for courts to concern themselves with perceptions of justice.

Optimal timing of partial outsourcing decisions

This article combines a real options approach to the optimal timing of outsourcing decisions with a linear programming technique for solving one-dimensional optimal stopping problems. We adopt a partial outsourcing model proposed by Y. Moon (2010) which assumes profit flows to follow a geometric Brownian motion and explicitly takes into account the benets and costs of all eorts which a firm spends on the project prior to the outsourcing date. The problem of deciding when to outsource and how much efort to spend is solved when the underlying profit flows or index processes are modeled by general one-dimensional diffusions. Optimal outsourcing times are proved to be of threshold type, and sensitivity results regarding market volatility and other quantities are derived. The corresponding optimal stopping problems are reformulated in terms of finnite dimensional linear programs and nonlinear optimization problems. These reformulations are exploited to prove sensitivity results in a novel way. Specific management recommendations are provided.

Forest species recognition using macroscopic images

The recognition of forest species is a very challenging task that generally requires well-trained human specialists. However, few reach good accuracy in classification due to the time taken for their training; then they are not enough to meet the industry demands. Computer vision systems are a very interesting alternative for this case. The construction of a reliable classification system is not a trivial task, though. In the case of forest species, one must deal with the great intra-class variability and also the lack of a public available database for training and testing the classifiers. To cope with such a variability, in this work, we propose a two-level divide-and-conquer classification strategy where the image is first divided into several sub-images which are classified independently. In the lower level, all the decisions of the different classifiers, trained with different features, are combined through a fusion rule to generate a decision for the sub-image. The higher-level fusion combines all these partial decisions for the sub-images to produce a final decision. Besides the classification system we also extended our previous database, which now is composed of 41 species of Brazilian flora. It is available upon request for research purposes. A series of experiments show that the proposed strategy achieves compelling results. Compared to the best single classifier, which is a SVM trained with a texture-based feature set, the divide-and-conquer strategy improves the recognition rate in about 9 percentage points, while the mean improvement observed with SVMs trained on different descriptors was about 19 percentage points. The best recognition rate achieved in this work was 97.77 %.

A stochastic analysis of asynchronous demand in disassembly processes of remanufacturing systems

Disassembly is the most critical stage of remanufacturing activities. The condition of parts disassembled for reuse/remanufacturing display a high variance. Hence demand for different parts found in a core cannot always be satisfied by a single core. At this point the question is whether or not to partially disassemble the second core. In this study, we concentrate on quantifying the potential benefits of this decision in a remanufacturing environment by using a queueing network model with kanban control. The model is solved by an approximate method, and the impact of profitability on partial disassembly decisions is studied by numerical analysis.

Novel partial decision combining schemes for Rayleigh fading

ABSTRACTNew combining schemes for partial decisions using one‐bit and two‐bit information of the received samples are derived for Rayleigh fading channels. Compared with previous partial decision combining schemes, these new schemes have either better performances or simpler structures, or both. Simulation results show that some of the new schemes can achieve considerable performance gains over the conventional partial decision combiner in Rayleigh fading or log‐normal shadowed Rayleigh fading, whereas their structures are much simpler than the conventional partial decision combiner. This confirms the advantages of the newly designed combining schemes. Copyright © 2011 John Wiley & Sons, Ltd.

Distributed Partially Cooperative NMPC Under Limited Communication and Destabilizing Interconnections

Abstract In this paper, a new formulation of distributed and partially cooperative control under limited communication is proposed. Unlike many existing schemes, destabilizing interconnections are considered and partial load shading-like decisions may be potentially taken in order to maintain the overall system integrity. The example of power systems black-out can be viewed as the targeted context although the present paper gives the general framework regardless any specific application.

Combining Classifiers in the ROC-space for Off-line Signature Verification

In this work we present a strategy for off-line signature verification. It takes into account a writer-independent model which reduces the pattern recognition problem to a 2-class problem, hence, makes it possible to build robust signature verification systems even when few signatures per writer are available. Receiver Operating Characteristic (ROC) curves are used to improve the performance of the proposed system. The contribution of this paper is two-fold. First of all, we analyze the impacts of choosing differ- ent fusion strategies to combine the partial decisions yielded by the SVM classifiers. Then ROC produced by different classifiers are combined using maximum likelihood analysis, producing an ROC combined classifier. Through comprehensive experiments on a database composed of 100 writers, we demonstrate that the ROC combined classifier based on the writer-independent approach can reduce considerably false rejection rate while keeping false acceptance rates at acceptable levels.

A narwhal is an animal: partial word knowledge biases adults' decisions.

Adults' accuracy on a sentence verification task was measured to assess reliable biases in their responses as a function of their level of knowledge about the target words, as well as adults' ability to reject incorrect ontological category membership for their partially known words. Three levels of word knowledge (known, frontier, and unknown) were assessed in two experiments. Results indicated that participants were conservative decision-makers about sentences using the vocabulary words and that the conservative bias increased in strength as their level of knowledge about the target word increased. When biased guessing was discouraged, results indicated that participants were able to accurately identify both correct and incorrect sentences using targets at all three levels of knowledge, including their partially known words, but were least reliable at rejecting false sentences using unknown targets. Implications for the role of "knowing not" in adults' vocabulary acquisition are discussed.

ESTIMATION OF ALTERNATIVE WALL INSULATING SOLUTIONS USING THE METHOD OF TECHNOLOGICAL NETWORK MODEL “CUTTING OUT” JUNCTIONS/SIENŲ ŠILTINIMO ALTERNATYVIŲ SPRENDIMŲ ĮVERTINIMAS, TAIKANT TECHNOLOGINIO TINKLINIO MODELIO MAZGŲ “IŠPJOVIMO” METODĄ

The article investigates the problem of the dwelling house wall insulation of many-variant decisions estimation. At present multi-criteria evaluation methods are used. They provide the formal algorithm of the final effective selection of solution. However, during the optimisation of the decisions it is expedient to analyse the efficiency of intermediate decisions and those influencing the final findings. With that end in view the author suggests the multi- criteria problem of the evaluations of the variant decision using the method or technological network model of “cutting out” junctions. The main points of this method are: To cut out a comparable variant junction. Comparable variant junction is such a junction that contains at least two technological process variants. To determine comparable partial processes to the junction. To carry out a multi-criteria evaluation. To eliminate non-rational variants from further calculations. To pick out the most effective decision for the junction. From such decisions the complex process variants of wall insulation are made and multi-criteria evaluation is repeated. This method increases the abilities to raise the number of the partial alternative decisions for establishing the effectiveness ot the intermediate decisions, to receive and verify quickly final results of the decision evaluation and to decrease the amount of the calculations. “Cutting out” junction method can be used for estimating the decisions of building thermal renovation at the designing stage.

Branch and bound crossed with GA to solve hybrid flowshops

This article deals with an optimal methods for solving a k-stage hybrid flowshop scheduling problem. This problem is known to be NP-hard. In 1991, Brah and Hunsucker proposed a branch and bound algorithm to solve this problem. However, for some medium size problems, the computation time is not acceptable. The aim of this article is to present an improvement of this algorithm. As a matter of fact, we prove that the value of their lower bound (LB) may decrease along a path of the search tree. First of all, we present an improvement of their LB. Then, we introduce several heuristics at the beginning of the search in order to compute an initial upper bound and genetic algorithm (GA) to improve during the search the value of the upper bound. More precisely, our GA takes into account the set of partial decisions made by the branch and bound and builds a series of populations of complete solutions with the aim of improving the upper bound (the best found criterion value corresponding to a complete solution). Experimentation show that the optimality of branch and bound is more often reached and the value of criterion is improved when our improvements are taken into account.

Performance comparison of two suboptimum detection procedures in real environment

Two suboptimum detection procedures, one based on the binary matched filtering and the other based on the weighting of partial decisions, were introduced earlier into the literature as very suitable for the cost-effective microprocessor-based implementation of the medium-speed voice-band data modems. Previous analysis of these two detection procedures was limited to the case when only Gaussian noise is present in the communications system. Here, both procedures are treated comparatively under conditions typical of realistic applications. These conditions refer to the cases when Gaussian noise is combined with intersymbol interference or timing jitter. To provide a reference against which the new detection procedures can be evaluated, results known from the literature for the optimal detector are here utilised and are comparatively presented.