Low-level Constraints Research Articles

A Knee Point Driven Evolutionary Algorithm for Multiobjective Bilevel Optimization.

Bilevel optimization is a special type of optimization in which one problem is embedded within another. The bilevel optimization problem (BLOP) of which both levels are multiobjective functions is usually called the multiobjective BLOP (MBLOP). The expensive computation and nested features make it challenging to solve. Most existing studies look for complete lower-level solutions for every upper-level variable. However, not every lower-level solution will participate in the bilevel Pareto-optimal front. Under a limited computational budget, instead of wasting resources to find complete lower-level solutions that may not be in the feasible region or inducible region of the MBLOP, it is better to concentrate on finding the solutions with better performance. Bearing these considerations in mind, we propose a multiobjective bilevel optimization solving routine combined with a knee point driven algorithm. Specifically, the proposed algorithm aims to quickly find feasible solutions considering the lower-level constraints in the first stage and then concentrates the computational resources on finding solutions with better performance. Besides, we develop several multiobjective bilevel test problems with different properties, such as scalable, deceptive, convexity, and (dis)continuous. Finally, the performance of the algorithm is validated on a practical petroleum refining bilevel problem, which involves a multiobjective environmental regulation problem and a petroleum refining operational problem. Comprehensive experiments fully demonstrate the effectiveness of our presented algorithm in solving MBLOPs.

UEFSD: Unsupervised medical images fusion based on exclusive features and saliency detection for SPECT-MRI images of glioma

Background and ObjectiveTo address the key feature extraction problem in glioma image fusion due to blurred edge and texture features, with certain exclusive features entangled with noise, we propose an unsupervised fusion network based on exclusive feature unentanglement and saliency detection (UEFSD). Single-photon emission computed tomography (SPECT) and magnetic resonance imaging (MRI) are taken as examples. MethodsFirst, the SPECT image in RGB space is converted to YCbCr space. Then, Y-channel SPECT and MRI images are fed into the UEFSD-based network and the attribute and object encoders are respectively used to separate the attribute vector of exclusive information for each modality and object feature mapping based on this attribute vector. Subsequently, the attribute vector and object feature maps are fused through the fusion decoder to obtain the fused image. In the training phase, first, the image is reconstructed using the attribute vector and object feature maps of the other modalities. The feature displaying the largest difference between the source and reconstructed images is identified as the exclusive feature that cannot be obtained via reconstruction. We trained the encoders and decoders in an unsupervised manner using exclusive feature loss as a high-level constraint and introducing a saliency mask as a low-level constraint. The trained attribute encoders and object encoders feed information to the fusion decoder to generate fused images. ResultUEFSD-based network was trained on Harvard Medical School's public dataset and tested by fusing MRI-T2 glioma images with their SPECT-T1 and SPECT-Tc counterparts. The experimental results demonstrate that UEFSD performs optimally in both subjective and objective indices when compared with nine other existing networks. ConclusionThe proposed UEFSD-based network can obtain clear edge and texture glioma features, which can effectively improve the rich and natural colour information of medical images and has the potential for computer-aided diagnosis system application.

Land Use and Land Cover Mapping in China Using Multimodal Fine-Grained Dual Network

With the advancement of geo-systems and the increased availability of satellite data, a plethora of Land-Use and Land-Cover (LULC) products have been developed. The existing LULC products primarily relied on time-series imagery to classify land by pixel-based classifiers, allowing for local analysis and accurate boundary detection. However, the advent of deep learning has shifted towards the use of patch-based CNN models for generating land cover maps. In this paper, (1) we create a training dataset for China using a voting strategy based on three off-the-shelf available LULC products, avoiding the labor-intensive manual annotation. (2) We design a novel CNN-based model for LULC task, called Multi-modal Fine-grained Dual Network (dubbed as Dual-Net), which takes dual-date images to generate final maps, and reduces the need for gap-free temporal sequences or separate cloud detection. To leverage the correlation between location, date, and category, we embed multi-modal information (dates and geo-locations) to the model. Further, by incorporating low-level constraints and using pseudo-label refinement, we continually improve the performance and achieve more refined segmentation. (3) Due to the lack of a suitable validation dataset for China, we create a new validation dataset called China Sentinel2 Validation Dataset (CSVD) by manually annotating 733 finely labeled images of 1024 × 1024 pixels of China-specific Sentinel2 data. (4) Extensive experiments demonstrate that our model outperforms existing LULC products and produces more fine-grained segmentation results comparable to other patch-based products. Finally, we release annual LULC maps for China in 2020-2022 and also make our model accessible online for real-time results export.

Optimality Conditions for Bilevel Imaging Learning Problems with Total Variation Regularization

We address the problem of optimal scale-dependent parameter learning in total variation image denoising. Such problems are formulated as bilevel optimization instances with total variation denoising problems as lower-level constraints. For the bilevel problem, we are able to derive M-stationarity conditions, after characterizing the corresponding Mordukhovich generalized normal cone and verifying suitable constraint qualification conditions. We also derive B-stationarity conditions, after investigating the Lipschitz continuity and directional differentiability of the lower-level solution operator. A characterization of the Bouligand subdifferential of the solution mapping, by means of a properly defined linear system, is provided as well. Based on this characterization, we propose a two-phase non-smooth trust-region algorithm for the numerical solution of the bilevel problem and test it computationally for two particular experimental settings.

Multi-Level Dependent-Chance Model for Hydropower Reservoir Operations

Some hydropower reservoirs are operated under different constraint levels. For these reservoirs, a multi-level (ML) dependent-chance (DC) model is established. In the model, only when the higher-level constraints are satisfied are the lower-level constraints or system benefits considered. The multi-level dependent-chance (MLDC) model is specified by two models. One is based on existing reliability-constrained (RC) dynamic programming (DP), in which the soft constraints are addressed using reliability constraints of 1, and the priorities are reflected using the order of magnitudes of Lagrange multipliers. The other is the explicit dependent-chance reasoning in the DP recursive function, in which each soft constraint is represented as an objective function of negative expected failure time and the optimum is the solution with a larger value for all higher-level objective functions. The proposed models are applied to derive long-term operation rules for the hydropower system on the middle-lower Lancang River. The results show the feasibility and performances of the explicit graded constraint control of the proposed model and the solution methods.

An exact projection-based algorithm for bilevel mixed-integer problems with nonlinearities

We propose an exact global solution method for bilevel mixed-integer optimization problems with lower-level integer variables and including nonlinear terms such as, e.g., products of upper-level and lower-level variables. Problems of this type are extremely challenging as a single-level reformulation suitable for off-the-shelf solvers is not available in general. In order to solve these problems to global optimality, we enhance an approximative projection-based algorithm for mixed-integer linear bilevel programming problems from the literature to become exact under one additional assumption. This assumption still allows for discrete and continuous leader and follower variables on both levels, but forbids continuous upper-level variables to appear in lower-level constraints and thus ensures that a bilevel optimum is attained. In addition, we extend our exact algorithm to make it applicable to a wider problem class. This setting allows nonlinear constraints and objective functions on both levels under certain assumptions, but still requires that the lower-level problem is convex in its continuous variables. We also discuss computational experiments on modified library instances.

Semismooth Newton-type method for bilevel optimization: global convergence and extensive numerical experiments

We consider the standard optimistic bilevel optimization problem, in particular upper- and lower-level constraints can be coupled. By means of the lower-level value function, the problem is transformed into a single-level optimization problem with a penalization of the value function constraint. For treating the latter problem, we develop a framework that does not rely on the direct computation of the lower-level value function or its derivatives. For each penalty parameter, the framework leads to a semismooth system of equations. This allows us to extend the semismooth Newton method to bilevel optimization. Besides global convergence properties of the method, we focus on achieving local superlinear convergence to a solution of the semismooth system. To this end, we formulate an appropriate CD-regularity assumption and derive sufficient conditions so that it is fulfilled. Moreover, we develop conditions to guarantee that a solution of the semismooth system is a local solution of the bilevel optimization problem. Extensive numerical experiments on 124 examples of nonlinear bilevel optimization problems from the literature show that this approach exhibits a remarkable performance, where only a few penalty parameters need to be considered.

Connecting Biological Detail With Neural Computation: Application to the Cerebellar Granule–Golgi Microcircuit

Neurophysiology and neuroanatomy constrain the set of possible computations that can be performed in a brain circuit. While detailed data on brain microcircuits is sometimes available, cognitive modelers are seldom in a position to take these constraints into account. One reason for this is the intrinsic complexity of accounting for biological mechanisms when describing cognitive function. In this paper, we present multiple extensions to the neural engineering framework (NEF), which simplify the integration of low-level constraints such as Dale's principle and spatially constrained connectivity into high-level, functional models. We focus on a model of eyeblink conditioning in the cerebellum, and, in particular, on systematically constructing temporal representations in the recurrent granule-Golgi microcircuit. We analyze how biological constraints impact these representations and demonstrate that our overall model is capable of reproducing key properties of eyeblink conditioning. Furthermore, since our techniques facilitate variation of neurophysiological parameters, we gain insights into why certain neurophysiological parameters may be as observed in nature. While eyeblink conditioning is a somewhat primitive form of learning, we argue that the same methods apply for more cognitive models as well. We implemented our extensions to the NEF in an open-source software library named "NengoBio" and hope that this work inspires similar attempts to bridge low-level biological detail and high-levelfunction.

Расширение представления о механизмах инсайтного решения в рамках теории изменения репрезентации С. Олссона

The paper explores an option to expand the model of mechanisms of insight problem solving proposed by S. Ohlsson. The proposed mechanisms of insight problem solving — chunk decomposition and constraint relaxation — considered within the framework of high-level and low-level processes. Chunk decomposition described as low-level mechanism and constraint relaxation as high-level mechanism. We assume that difficulty of the different insight problems can be explained by high level chunk decomposition and low-level constraint relaxation. The paper describes two experiments dedicated to verify the assumption. The first experiment examines process of solving anagrams (with word) as high-level chunk decomposition. The main results of the experiment show the prospect of distinguishing the semantic chunk decomposition as insight problem solving mechanism. The second experiment use nine-dot insight problem to examine its solution process as relaxation of low-level (perceptual) constraints. Based on the results of the experiments, it can be considered possible to conclude that the expansion of S. Ohlsson’s model of mechanisms of insight problem solving allows to deem the solutions of various problems in a unified system.

Challenges and Opportunities with New Generation Geostationary Meteorological Satellite Datasets for Analyses and Initial Conditions for Forecasting Hurricane Irma (2017) Rapid Intensification Event

This study utilizes an extremely high spatial resolution GOES-16 atmospheric motion vector (AMV) dataset processed at 15 min intervals in a modified version of our original dynamic initialization technique to analyze and forecast a rapid intensification (RI) event in Hurricane Irma (2017). The most important modifications are a more time-efficient dynamic initialization technique and adding a near-surface wind field adjustment as a low-level constraint on the distribution of deep convection relative to the translating center. With the new technique, the Coupled Ocean/Atmospheric Mesoscale Prediction System for Tropical Cyclones (COAMPS-TC) model initial wind field at 12.86 km elevation quickly adjusts to the cirrus-level GOES-16 AMVs to better detect the Irma outflow magnitude and areal extent every 15 min, and predicts direct connections to adjacent synoptic circulations much better than a dynamic initialization with only lower-resolution hourly GOES-13 AMVs and also better than a cold-start COAMPS-TC initialization with a bogus vortex. Furthermore, only with the GOES-16 AMVs does the COAMPS-TC model accurately predict the timing of an intermediate 12 h constant-intensity period between two segments of the Irma RI. By comparison, HWRF model study of the Irma case that utilized the same GOES-16 AMV dataset predicted a continuous RI without the intermediate constant-intensity period, and predicted more limited outflow areal extents without strong direct connections with adjacent synoptic circulations.

Pyramidal Temporal Pooling With Discriminative Mapping for Audio Classification

Audio signals are temporally-structured data, and learning their discriminative representations containing temporal information is crucial for the audio classification. In this article, we propose an audio representation learning method with a hierarchical pyramid structure called pyramidal temporal pooling (PTP) which aims to capture the temporal information of an entire audio sample. By stacking a global temporal pooling layer on multiple local temporal pooling layers, the PTP can capture the high-level temporal dynamics of the input feature sequence in an unsupervised way. Furthermore, in the top global temporal pooling layer, we jointly optimize a learnable discriminative mapping (DM) and a softmax classifier. Such that, a joint learning method for the discriminative audio representations and the classifier called DM-PTP is also presented. By treating the temporal encoding as a low-level constraint of a bi-level optimization problem, the DM-PTP can produce the discriminative representation while maintaining the temporal information of the whole sequence. For an audio sample with an arbitrary time duration, both our PTP and DM-PTP can encode the input feature sequence with arbitrary length into a fixed-length representation. Without using any data augmentation and ensemble learning methods, both PTP and DM-PTP outperform the state-of-the-art CNNs on the audio event recognition (AER) dataset, and can achieve comparable performance on the DCASE 2018 acoustic scene classification (ASC) dataset compared with other best models in the challenge.

Structured variation in Suzhou Chinese fricative consonants and vowels

Uniform phonetic implementation of phonological primitives has been suggested to be a low-level constraint on speech production. Suzhou Chinese offers an example of how this constraint may shape the development of systems of phonological contrast: a pair of recently innovated “fricative vowels,” rounded and unrounded, have come to be produced with the addition of noticeable fricative noise. This noise is spectrally similar to that of a fricative consonant, most frequently alveolopalatal /ɕ/ but occasionally /s/, suggesting that speakers have begun to refer to shared acoustic targets for these phonotactically dissimilar sounds. In order to assess the degree of acoustic similarity among these segments, we compare the fricative noise in the unrounded and rounded fricative vowels of 43 speakers of Suzhou Chinese with the fricative noise of each speaker’s onset fricative consonants /ɕ/ and /s/. Spectral parameters analyzed are chosen to reflect constriction anteriority in strident fricatives. Correlations in these parameters as implemented across consonantal onsets such as /ɕ/ and /s/, and the unrounded fricative vowel and the rounded fricatie vowel are examined.

Hierarchical Control with Guaranteed Fault Diagnosability

A multiple time-scale hierarchical control approach with guarantees for active fault diagnosability (detection and isolation) is presented. Robust positive invariance notions are used to provide a set-membership test for FDI validation at the low level and sufficient constraints for guaranteeing FDI at the high level. Due to the computational burden of the involved bilinear optimization problem, the high-level reference governor operates at a slower sampling time than the controller of the low-level system. Therefore, corner cutting avoidance constraints are imposed such that the faster-sampled lower level is still guaranteed to function properly. A case study illustrates the presented theoretical notions.

Static compliance checking beyond separation of duty constraints

In many application domains, processes often have to comply with business level rules and policies, called compliance constraint. Separation of duty requirement is one of such constraint, and enhancement measures are taken to make sure the business process is compliance with the constraints. Separation of duty Algebra is a high level policy designed to specify Separation of duty constraints, which can describe both quantification requirements and qualification requirements. In this article, we improve the high level description SoDA by involving task information. The improved specification expresses relations between user and task, describes bind of duty constraint beyond separation of duty. In order to make the high level policy act on concrete process, we proceed by constructing mapping rules to translate high level policies to low level constraints, which is representing by Description Logic. Then, we propose a framework for business process compliance checking. At last we give a case study to show the suitably of the method.

Interpersonal Movement Synchrony Responds to High- and Low-Level Conversational Constraints.

Much work on communication and joint action conceptualizes interaction as a dynamical system. Under this view, dynamic properties of interaction should be shaped by the context in which the interaction is taking place. Here we explore interpersonal movement coordination or synchrony—the degree to which individuals move in similar ways over time—as one such context-sensitive property. Studies of coordination have typically investigated how these dynamics are influenced by either high-level constraints (i.e., slow-changing factors) or low-level constraints (i.e., fast-changing factors like movement). Focusing on nonverbal communication behaviors during naturalistic conversation, we analyzed how interacting participants' head movement dynamics were shaped simultaneously by high-level constraints (i.e., conversation type; friendly conversations vs. arguments) and low-level constraints (i.e., perceptual stimuli; non-informative visual stimuli vs. informative visual stimuli). We found that high- and low-level constraints interacted non-additively to affect interpersonal movement dynamics, highlighting the context sensitivity of interaction and supporting the view of joint action as a complex adaptive system.

School Absenteeism and Child Labor in Rural Bangladesh

Evidence from many countries suggests a close association between school absenteeism and high rates of school drop-out, grade repetition and low academic achievement, and health and psychological disorders. In rural Bangladesh, school absenteeism is a major issue. Many children who are enrolled do not attend school regularly. Our own calculation using the Multiple Indicator Survey (MICS) 3 2005-06 data shows that more than a quarter of children aged 7-14 years missed at least one day of school in a six-day school week in the rural areas of Bangladesh. In addition, Bangladesh has a very high incidence of child labor. In 2005-06, 41% of children aged 7-14 years participated in the labor market activities in the rural areas. In this paper, we empirically examine the major determinants of school absenteeism and child labor among children aged 7-14 years and number of hours worked by them in rural Bangladesh using the MICS 3 2005-06 data. Since, data on school absenteeism and number of hours worked are censored, we estimate a bivariate Tobit model. Our empirical analysis shows that parental educational level has a significant negative effect on both school absenteeism and number of hours worked by children. The negative effect of education level of mother on school absenteeism is twice as large as that of father. We find that children belonging to poorer households have higher level of school absenteeism and work greater number of hours. Our results suggest that low income and financial constraints continue to act as important barriers to improving schooling outcomes in rural Bangladesh. We also find evidence of gender bias against female children regarding child labor. Female children work significantly greater number of hours than male children. In addition, the gender differential in number of hours worked increases with age. Finally, we find that school absenteeism is lower among children enrolled in higher grades. Results show that low parental education level and significant resource constraints both in terms of wealth and time are main obstacles in improving schooling outcomes for children in rural Bangladesh. There is continuing need to provide financial support to poor households and increase their access to credit market. Reducing school absenteeism in lower grades can help in reducing school drop-outs and improving academic achievement. There is need to target government assistance and aid programs towards children in lower grades and their parents.

Following control approach based on model predictive control for wheeled inverted pendulum robot

Personal robots, which are seen as tools that will be needed to support our aging society, will be expected to support the comfortable lifestyles of healthy young people as well as the elderly. However, excessive and premature robot support may adversely impact the physical abilities of their human owner/operators. In this paper, the authors propose a personal robot equipped with wheeled inverted pendulum control that can carry baggage and follow the human being. Since such robots could remove the drudgery associated with carrying luggage, their use could also encourage people to go outside and walk briskly, which could contribute to improved health management. This paper proposes a novel control approach for a robot following the human being. The proposed approach employs a model predictive control that facilitates consideration of several types of upper and lower level constraints a personal robot would require. The effectiveness of our proposed approach was then verified in experiments using a prototype personal robot.

One more year of schooling or work: Grade-transition and child labor in rural Bangladesh

This paper aims to identify the main factors hindering the achievement of universal primary education (Millennium Development Goal 2) for children in Rural Bangladesh using a large household survey data. It empirically examine factors affecting the grade-transition and the labor force participation of children aged 7-14 years in rural Bangladesh using the Multiple Indicator Cluster Survey (MICS) 3 2005-2006 data. It addresses following important questions: (i) effects of family background and resources and child-specific characteristics on the grade-transition and child labor and (ii) whether these effects vary across grade levels. It uses the bivariate probit model to jointly estimate the grade-transition and the labor force participation equations. Results show that parental education has a significant positive effect on the grade-transition and a significant negative effect on the labor force participation. Children of more educated parents are more likely to transit to the next grade and just study and are less likely to participate in the labor force, combine study and work, and just work. It also finds that the positive effect of mother’s education on the probability of grade-transition and just study is larger on children in higher grades. On the other hand, it has a greater negative effect on the probability of just work on children in higher grades. Overall results show that father’s education has greater effect on the grade-transition in lower grades, while mother’s education has greater effect on the grade-transition in higher grades. Results suggest that birth-order has larger effect on the grade-transition and the labor force participation of children in lower grades. First-born children in lower grades are less likely to transit to the next grade and just study. On the other hand, they are more likely to participate in the labor force, combine study and work, and just work. Similarly, results show that children who experienced interruptions in schooling in lower grades are more likely to participate in the labor force and combine study and work, and are less likely to just study compared to children who experienced interruptions in schooling in higher grades. Results suggest that low parental education level and significant resource constraints both in terms of wealth and time are main obstacles in achieving the goal of universal primary education for children. Many children lack adequate learning environment at home, in part due to low parental educational attainment. There is an urgent need to focus on increasing the capacity, financial and non-financial, of rural households. Conditional cash transfers, micro-finance programs, interest-free educational loans, and scholarships, particularly targeted towards poorer households and female students, will enhance educational investments and reduce child labor.

Low level constraints on dynamic contour path integration.

Contour integration is a fundamental visual process. The constraints on integrating discrete contour elements and the associated neural mechanisms have typically been investigated using static contour paths. However, in our dynamic natural environment objects and scenes vary over space and time. With the aim of investigating the parameters affecting spatiotemporal contour path integration, we measured human contrast detection performance of a briefly presented foveal target embedded in dynamic collinear stimulus sequences (comprising five short ‘predictor’ bars appearing consecutively towards the fovea, followed by the ‘target’ bar) in four experiments. The data showed that participants' target detection performance was relatively unchanged when individual contour elements were separated by up to 2° spatial gap or 200 ms temporal gap. Randomising the luminance contrast or colour of the predictors, on the other hand, had similar detrimental effect on grouping dynamic contour path and subsequent target detection performance. Randomising the orientation of the predictors reduced target detection performance greater than introducing misalignment relative to the contour path. The results suggest that the visual system integrates dynamic path elements to bias target detection even when the continuity of path is disrupted in terms of spatial (2°), temporal (200 ms), colour (over 10 colours) and luminance (−25% to 25%) information. We discuss how the findings can be largely reconciled within the functioning of V1 horizontal connections.

A greedy gradient-simulated annealing selection hyper-heuristic

Educational timetabling problem is a challenging real world problem which has been of interest to many researchers and practitioners. There are many variants of this problem which mainly require scheduling of events and resources under various constraints. In this study, a curriculum based course timetabling problem at Yeditepe University is described and an iterative selection hyper-heuristic is presented as a solution method. A selection hyper-heuristic as a high level methodology operates on the space formed by a fixed set of low level heuristics which operate directly on the space of solutions. The move acceptance and heuristic selection methods are the main components of a selection hyper-heuristic. The proposed hyper-heuristic in this study combines a simulated annealing move acceptance method with a learning heuristic selection method and manages a set of low level constraint oriented heuristics. A key goal in hyper-heuristic research is to build low cost methods which are general and can be reused on unseen problem instances as well as other problem domains desirably with no additional human expert intervention. Hence, the proposed method is additionally applied to a high school timetabling problem, as well as six other problem domains from a hyper-heuristic benchmark to test its level of generality. The empirical results show that our easy-to-implement hyper-heuristic is effective in solving the Yeditepe course timetabling problem. Moreover, being sufficiently general, it delivers a reasonable performance across different problem domains.