General Instances Research Articles

The electric on-demand bus routing problem with partial charging and nonlinear function

The electric vehicle routing problems (EVRPs) with recharging policy consider the limited range of electric vehicles and thus include intermediate visits to charging stations (CSs). In general, minimizing the resultant charging costs such as charging duration or charging amount is also part of the objective of the EVRP. Accordingly, the EVRP has received considerable attention over the past years. Nevertheless, this type of problem in the domain of passenger transportation, a VRP variant, has been rarely studied in the literature, especially with time windows, a realistic nonlinear charging function or a partial charging policy. Hence this research extends the existing work on the EVRP to the on-demand bus routing problem (ODBRP) which transports passengers with the bus station assignment (BSA). The resultant problem is the electric ODBRP (EODBRP). Specifically, each passenger can have more than one stations to board or alight, and they are assigned to the ones with the smallest increase in the total user ride time (URT). In the EODBRP, frequent intermediate visits to CSs are considered. Moreover, a nonlinear charging function is used and the partial charging strategy is applied. To solve the EODBRP, a greedy insertion method with a ‘charging first, routing second’ strategy is developed, followed by a large neighborhood search (LNS) which consists of local search (LS) operators to further improve the solution quality. Experimental data are generated by a realistic instance generator based on a real city map, and the corresponding results show that the proposed heuristic algorithm performs well in solving the EODBRP. Finally, sensitivity analyses with divergent parameters such as the temporal distributions of passengers and bus ranges may provide practical guidance.

The Aesthetics of Disharmony: Harnessing Sounds and Images for Dynamic Soundscapes Generation

This work presents an autonomous approach that explores the dynamic generation of relaxing soundscapes for games and artistic installations. Differently from past works, this system can generate music and images simultaneously, preserving human intent and coherency. We present our algorithm for the generation of audiovisual instances and also a system based on this approach, verifying the quality of the outcomes it can produce in light of current approaches for the generation of images and music. We also instigate the discussion around the new paradigm in arts, where the creative process is delegated to autonomous systems, with limited human participation. Our user study (N=74) shows that our approach overcomes current deep learning models in terms of quality, being recognized as human production, as if the outcome were being generated out of an endless musical improvisation performance.

GBOML: a structure-exploiting optimization modelling language in Python

Mixed-Integer Linear Programs (MILPs) have many practical applications. Most modelling tools for MILPs fall in two broad categories. Indeed, tools such as algebraic modelling languages allow practitioners to compactly encode models using syntax close to mathematical notation but usually lack support for special structures, while other tools instead provide predefined components that can be easily assembled but modifying or adding new components is difficult. In this work, we present the inner workings of the Graph-Based Optimization Modelling Language (GBOML), an open-source modelling tool implemented in Python combining the strengths of both worlds. GBOML natively supports special structures that can be encoded by a hierarchical hypergraph, offers syntax close to mathematical notation and facilitates the modular construction and reuse of time-indexed models. We detail design choices enabling these features and show that they simplify problem encoding, lead to faster instance generation times and sometimes faster solve times. We benchmark the times taken by GBOML, JuMP, Plasmo, Pyomo and AMPL to generate instances of a structured MILP. We find that GBOML outperforms Plasmo and Pyomo, is tied with JuMP but is slower than AMPL. With parallel model generation, GBOML outperforms JuMP and closes the gap with AMPL. GBOML has the smallest memory footprint.

Learning Context Restrained Correlation Tracking Filters via Adversarial Negative Instance Generation.

The tracking performance of discriminative correlation filters (DCFs) is often subject to unwanted boundary effects. Many attempts have already been made to address the above issue by enlarging searching regions over the last years. However, introducing excessive background information makes the discriminative filter prone to learn from the surrounding context rather than the target. In this article, we propose a novel context restrained correlation tracking filter (CRCTF) that can effectively suppress background interference via incorporating high-quality adversarial generative negative instances. Concretely, we first construct an adversarial context generation network to simulate the central target area with surrounding background information at the initial frame. Then, we suggest a coarse background estimation network to accelerate the background generation in subsequent frames. By introducing a suppression convolution term, we utilize generative background patches to reformulate the original ridge regression objective through circulant property of correlation and a cropping operator. Finally, our tracking filter is efficiently solved by the alternating direction method of multipliers (ADMM). CRCTF demonstrates the accuracy performance on par with several well-established and highly optimized baselines on multiple challenging tracking datasets, verifying the effectiveness of our proposed approach.

Few-Shot Pixel-Precise Document Layout Segmentation via Dynamic Instance Generation and Local Thresholding.

Over the years, the humanities community has increasingly requested the creation of artificial intelligence frameworks to help the study of cultural heritage. Document Layout segmentation, which aims at identifying the different structural components of a document page, is a particularly interesting task connected to this trend, specifically when it comes to handwritten texts. While there are many effective approaches to this problem, they all rely on large amounts of data for the training of the underlying models, which is rarely possible in a real-world scenario, as the process of producing the ground truth segmentation task with the required precision to the pixel level is a very time-consuming task and often requires a certain degree of domain knowledge regarding the documents at hand. For this reason, in this paper, we propose an effective few-shot learning framework for document layout segmentation relying on two novel components, namely a dynamic instance generation and a segmentation refinement module. This approach is able of achieving performances comparable to the current state of the art on the popular Diva-HisDB dataset, while relying on just a fraction of the available data.

Hybrid differential evolution algorithm for the resource constrained project scheduling problem with a flexible project structure and consumption and production of resources

The resource constrained project scheduling problem with a flexible project structure and consumption and production of resources, involves making a selection of activities and scheduling these activities in order to minimize the makespan, subject to precedence and resource constraints. Since finding a feasible selection of activities is NP-hard, we introduce the concept of group graphs and restrict ourselves to instances with an acyclic group graph. For these instances, which represent many practical cases, we show how to make a feasible selection of activities in polynomial time and use this concept to schedule the selected activities using a hybrid differential evolution algorithm. We compare this algorithm with an algorithm from the literature on special cases of instances without consumption and production of resources, and show that our algorithm creates solutions of higher quality. Furthermore, to compare general instances, we develop an ant colony optimization algorithm that performs slightly better on special cases than the algorithm from literature and show that the hybrid differential evolution algorithm outperforms the ant colony optimization algorithm on general instances.

Few-shot biomedical named entity recognition via knowledge-guided instance generation and prompt contrastive learning.

Few-shot learning (FSL) that can effectively perform named entity recognition in low-resource scenarios has raised growing attention, but it has not been widely studied yet in the biomedical field. In contrast to high-resource domains, biomedical named entity recognition (BioNER) often encounters limited human-labeled data in real-world scenarios, leading to poor generalization performance when training only a few labeled instances. Recent approaches either leverage cross-domain high-resource data or fine-tune the pre-trained masked language model using limited labeled samples to generate new synthetic data, which is easily stuck in domain shift problems or yields low-quality synthetic data. Therefore, in this paper, we study a more realistic scenario, i.e., few-shot learning for BioNER. Leveraging the domain knowledge graph, we propose knowledge-guided instance generation for few-shot BioNER, which generates diverse and novel entities based on similar semantic relations of neighbor nodes. In addition, by introducing question prompt, we cast BioNER as question answering (QA) task and propose prompt contrastive learning to improve the robustness of the model by measuring the mutual information (MI) between query-answer pairs. Extensive experiments conducted on various few-shot settings show that the proposed framework achieves superior performance. Particularly, in a low-resource scenario with only 20 samples, our approach substantially outperforms recent state-of-the-art (SoTA) models on four benchmark datasets, achieving an average improvement of up to 7.1% F1. Our source code and data are available at https://github.com/cpmss521/KGPC. Supplementary data are available at Bioinformatics online.

Positive unlabeled learning with tensor networks

Positive unlabeled learning is a binary classification problem with positive and unlabeled data. It is common in domains where negative labels are costly or impossible to obtain, e.g., medicine and personalized advertising. Most approaches to positive unlabeled learning apply to specific data types (e.g., images, categorical data) and can not generate new positive and negative samples. This work introduces a feature-space distance-based tensor network approach to the positive unlabeled learning problem. The presented method is not domain specific and significantly improves the state-of-the-art results on the MNIST image and 15 categorical/mixed datasets. The trained tensor network model is also a generative model and enables the generation of new positive and negative instances.

Constructing a Novel E-Learning Course, Educational Computational Chemistry through Instructional Design Approach in the TPASK Framework

The educational scenario after the COVID-19 confinement presents new challenges for teachers. Technological advances require teachers to be prepared for instruction through technology, and with this, the need for e-learning courses arose to strengthen this knowledge. This article aims to describe an innovative e-learning course in Educational Computational Chemistry (ECC) for in-service chemistry teachers through an Instructional Design (ID) that allows the development of the constructs associated with the Technological Pedagogical Science Knowledge (TPASK) framework. From the literature overview, relevant findings were raised concerning ID and its potential technological support. The results indicate that an effective ID must present general elements, such as the organisation and generation of content, progress monitoring, and feedback instances. However, the stages of engagement, flexibility, and positioning are relevant elements. These design elements are linked to emerging technological tools, such as artificial intelligence for generating audiovisual material, interactive content development, and event logs. In addition, positive results are evident from the teachers who participated in the ECC e-learning course, who project the knowledge, computer skills, and learning acquired into their professional work as chemistry teachers. Based on the above, a course design for ECC is proposed with general guidelines that contribute to the continuous training of in-service chemistry teachers.

A multi-objective mixed integer linear programming model for thesis defence scheduling

In this paper, we address the thesis defence scheduling problem, a critical academic scheduling management process, which has been overshadowed in the literature by its counterparts, course timetabling and exam scheduling. Specifically, we address the single defence assignment type of thesis defence scheduling problems, where each committee is assigned to a single defence, scheduled for a specific day, hour and room. We formulate a multi-objective mixed-integer linear programming model, which aims to be applicable to a broader set of cases than other single defence assignment models present in the literature, which have a focus on the characteristics of their universities. For such a purpose, we introduce a different decision variable, propose constraint formulations that are not regulation and policy specific, and cover and offer new takes on the more common objectives seen in the literature. We also include new objective functions based on our experience with the problem at our university and by applying knowledge from other academic scheduling problems. We also propose a two-stage solution approach. The first stage is employed to find the number of schedulable defences, enabling the optimisation of instances with unschedulable defences. The second stage is an implementation of the augmented ϵ-constraint method, which allows for the search of a set of different and non-dominated solutions while skipping redundant iterations. The methodology is tested for case-studies from our university, significantly outperforming the solutions found by human schedulers. A novel instance generator for thesis scheduling problems is presented. Its main benefit is the generation of the availability of committee members and rooms in availability and unavailability blocks, resembling their real-world counterparts. A set of 96 randomly generated instances of varying sizes is solved and analysed regarding their relative computational performance, the number of schedulable defences and the distribution of the considered types of iterations. The proposed method can find the optimal number of schedulable defences and present non-dominated solutions within the set time limits for every tested instance.

Fake Product Detection Using Blockchain Technology

Abstract: There are several fake products in the existing supply chain markets. It is necessary to have a system for end users to check all details about the product they are buying so that the customer can check if the product is genuine. In recent years, faking products play an important role in product manufacturing industries. Faking products affects the company name, sales, and profit of the organizations. Blockchain technology is used for the identification of actual merchandise and the detecting of faux products. Blockchain technology is generally a tally system that stores all the data of the deals that take place in it. The precise aspect approximately this era is that the tally we referred to is a allotted tally in a peer-to- peer community. Blockchain generation is relaxed as the facts saved once within the chain is immutable consequently any block cannot be modified or hacked. with the aid of using Blockchain era, clients or customers do no longer want to rely on users for confirmation of product authenticity and protection. Our device affords the rising generation of web use instances, and brief reaction (QR) codes offer a unique but affectable method to triumph over these faking of products. Counterfeited merchandise may be detected the use of a QR code scanner, wherein the QR code of the product is linked to the Blockchain. So, this system can be used to store product info and generate particular code of that product as blocks inside the database. It collects the precise code from the consumer and compares the code against entities within the Blockchain database. If the code matches, it will give all the information about the product otherwise no data can be outputted to the purchaser which shows that the product is fake or counterfeited.

Multi-Class Fuzzy-LORE: A Method for Extracting Local and Counterfactual Explanations Using Fuzzy Decision Trees

Multi-class classification is a fundamental task in Machine Learning. However, complex models can be viewed as black boxes, making it difficult to gain insight into how the model makes its predictions and build trust in its decision-making process. This paper presents a novel method called Multi-Class Fuzzy-LORE (mcFuzzy-LORE) for explaining the decisions made by multi-class fuzzy-based classifiers such as Fuzzy Random Forests (FRF). mcFuzzy-LORE is an adaptation of the Fuzzy-LORE method that uses fuzzy decision trees as an alternative to classical decision trees, providing interpretable, human-readable rules that describe the reasoning behind the model’s decision for a specific input. The proposed method was evaluated on a private dataset that was used to train an FRF-based multi-class classifier that assesses the risk of developing diabetic retinopathy in diabetic patients. The results show that mcFuzzy-LORE outperforms prior classical LORE-based methods in the generation of counterfactual instances.

Automated generation of scientific workflow generators with WfChef

Scientific workflow applications have gained significant importance, and their automated and efficient execution on large-scale computing platforms has been the subject of extensive research and development. For these efforts to be successful, a solid experimental methodology is needed to evaluate workflow algorithms and systems. A foundation for this methodology is the availability of realistic workflow instances. Although public repositories provide workflow instances for a few scientific applications, these are limited in scope, and workflow instances are not available for all application scales of interest. To address this limitation, previous work has developed generators of synthetic workflow instances of arbitrary scales. Despite being popular, the implementation of these generators is a manual and labor-intensive process that requires expert application knowledge. As a result, these generators only target a handful of applications, even though there are hundreds of workflow applications in production.We introduce WfChef , a fully automated framework for constructing a synthetic workflow generator for any scientific application. Based on an input set of workflow instances for a particular application, WfChef automatically produces a synthetic workflow generator. To measure the realism of the generated workflows, we define and evaluate several metrics. Using these metrics, we compare the realism of the workflows generated by WfChef generators to that of the workflows generated by the previously available, hand-crafted generators. We find that WfChef generators not only require zero development effort (because they are automatically produced), but also generate workflows that are more realistic than those generated by hand-crafted generators.

Integer programming column generation: accelerating branch-and-price using a novel pricing scheme for finding high-quality solutions in set covering, packing, and partitioning problems

AbstractLarge-neighbourhood search (LNS) heuristics are important mathematical programming techniques that search for primal feasible solutions by solving an auxiliary problem with a restricted feasible region. Extending such powerful generic LNS heuristics to the branch-and-price context is inherently challenging. The most prominent challenges arise from the fact that in branch-and-price algorithms, columns are generated with the sole aim to solve linear programming relaxations. Hence, the ability to form integer feasible solutions is not considered during the generation of columns. Without any changes to the standard pricing schemes, the potential of deploying generic LNS heuristics within a branch-and-price procedure is severely limited. This paper proposes a matheuristic, based on an LNS heuristic framework, where the novelty is a customised pricing scheme for generating columns to solve an auxiliary problem. The theoretical foundation for this pricing scheme is a set of optimality conditions for integer programs. From this foundation, a column generation strategy is developed for finding columns that are likely to be of use in high-quality primal feasible solutions for the original problem. The proposed matheuristic is implemented in the generic branch-price-and-cut solver GCG. On a broad test set comprising classical block diagonal structured instances and general instances from the MIPLIB 2017 Collection, the computational results show a significant improvement to the solving performance of GCG.

On the generation of metric TSP instances with a large integrality gap by branch-and-cut

This paper introduces a computational method for generating metric Travelling Salesman Problem (TSP) instances having a large integrality gap. The method is based on the solution of an integer programming problem, called IH-OPT, that takes as input a fractional solution of the Subtour Elimination Problem (SEP) on a TSP instance and computes a TSP instance having an integrality gap larger than or equal to the integrality gap of the first instance. The decision variables of IH-OPT are the entries of the TSP cost matrix, and the constraints are defined by the intersection of the metric cone with an exponential number of inequalities, one for each possible TSP tour. Given the very large number of constraints, we have implemented a branch-and-cut algorithm for solving IH-OPT. Then, by sampling cost vectors over the metric polytope and by solving the corresponding SEP, we can generate random fractional vertices of the SEP polytope. If we solve the IH-OPT problem for every sampled vertex using our branch-and-cut algorithm, we can select the generated TSP instance (i.e., cost vector), yielding the longest runtime for Concorde, the state-of-the-art TSP solver. Our computational results show that our method is very effective in producing challenging instances. As a by-product, we release the Hard-TSPLIB, a library of 41 small metric TSP instances which have a large integrality gap and are challenging in terms of runtime for Concorde.

High‐fidelity modeling of the braiding process for generating the realistic mesostructure of preforms in braided composites

AbstractAn accurate method for modeling the braided preforms is vital for the mesoscopic performance analysis of tubular composites. This study proposed a high‐fidelity virtual braiding model for the general braiding process, which further considers the fiber‐level deformation of the yarns during the braiding process to generate the realistic mesostructure of preforms in tubular composites. The method uses the digital element approach previously proposed for weaving processes to model the braiding yarn. The whole braiding process, including the deposition zone and convergence zone, is incorporated into the model to account for the effect of interaction in these zones on the preform mesostructure. Two braiding processes were simulated with the proposed virtual braiding method. The simulated geometry of the braided preforms and the yarn paths in the convergence zone agree well with the reported experimental results. To facilitate the performance analysis of braided composite structures, further automated generation of solid cell instances based on the obtained braided preforms was studied. Importantly, this work could reduce the considerable effort required for characterizing the mesostructure via the costly micro‐CT scans for braided composites, and the generated geometry of the braided preform is more precise than the widely used highly idealized models.

Improving Active Learning Performance through the Use of Data Augmentation

Active learning (AL) is a well-known technique to optimize data usage in training, through the interactive selection of unlabeled observations, out of a large pool of unlabeled data, to be labeled by a supervisor. Its focus is to find the unlabeled observations that, once labeled, will maximize the informativeness of the training dataset, therefore reducing data-related costs. The literature describes several methods to improve the effectiveness of this process. Nonetheless, there is a paucity of research developed around the application of artificial data sources in AL, especially outside image classification or NLP. This paper proposes a new AL framework, which relies on the effective use of artificial data. It may be used with any classifier, generation mechanism, and data type and can be integrated with multiple other state-of-the-art AL contributions. This combination is expected to increase the ML classifier’s performance and reduce both the supervisor’s involvement and the amount of required labeled data at the expense of a marginal increase in computational time. The proposed method introduces a hyperparameter optimization component to improve the generation of artificial instances during the AL process as well as an uncertainty-based data generation mechanism. We compare the proposed method to the standard framework and an oversampling-based active learning method for more informed data generation in an AL context. The models’ performance was tested using four different classifiers, two AL-specific performance metrics, and three classification performance metrics over 15 different datasets. We demonstrated that the proposed framework, using data augmentation, significantly improved the performance of AL, both in terms of classification performance and data selection efficiency (all the codes and preprocessed data developed for this study are available at https://github.com/joaopfonseca/publications/).

The Hutong neighbourhood grammar: A procedural modelling approach to unravel the rationale of historical Beijing urban structure

Hutong neighbourhoods, composed of Chinese courtyard dwellings (Siheyuan), are historically and socially significant urban spaces that embody the traditional Chinese way of life and philosophy. As part of the national heritage, there is an increasing research interest in Hutong neighbourhoods, many of which are facing oblivion. This study presents a formal grammar for Hutong neighbourhood generation. This research investigates traditional principles of urban planning of ancient Beijing, based on examples on the historical map Qianlong Jingcheng Quantu, to derive the lost design rules. These rules are used to build up a procedural modelling framework, which reveals the development of Beijing's urban structure from the Yuan (1271–1368) to the Qing (1644–1911) dynasty. Our findings present a grammar incorporated into the procedural modelling framework to parametrically generate Hutong neighbourhoods, which replicates the morphological characteristics of historic cases. It contributes to the understanding of the generation of Hutong neighbourhoods. In support of heritage sustainability, this grammar can be implemented in a computational environment by visual scripting that enables the generation of new instances of Hutong neighbourhoods, both real and virtual.

Generating Random SAT Instances: Multiple Solutions could be Predefined and Deeply Hidden

The generation of SAT instances is an important issue in computer science, and it is useful for researchers to verify the effectiveness of SAT solvers. Addressing this issue could inspire researchers to propose new search strategies. SAT problems exist in various real-world applications, some of which have more than one solution. However, although several algorithms for generating random SAT instances have been proposed, few can be used to generate hard instances that have multiple predefined solutions. In this paper, we propose the KHidden-M algorithm to generate SAT instances with multiple predefined solutions that could be hard to solve by the local search strategy when the number of predefined solutions is small enough and the Hamming distance between them is not less than half of the solution length. Specifically, first, we generate an SAT instance that is satisfied by all of the predefined solutions. Next, if the generated SAT instance does not satisfy the hardness condition, then a strategy will be conducted to adjust clauses through multiple iterations to improve the hardness of the whole instance. We propose three strategies to generate the SAT instance in the first part. The first strategy is called the random strategy, which randomly generates clauses that are satisfied by all of the predefined solutions. The other two strategies are called the estimating strategy and greedy strategy, and using them, we attempt to generate an instance that directly satisfies or is closer to the hardness condition for the local search strategy. We employ two SAT solvers (i.e., WalkSAT and Kissat) to investigate the hardness of the SAT instances generated by our algorithm in the experiments. The experimental results show the effectiveness of the random, estimating and greedy strategies. Compared to the state-of-the-art algorithm for generating SAT instances with predefined solutions, namely, M-hidden, our algorithm could be more effective in generating hard SAT instances.

A LINDBLADIAN OPERATOR FOR OPEN SYSTEM THERMALIZATION, WITH APPLICATIONS

A general phenomenological expression is provided, within the frame of the Gorini-Kossakowski-Sudarshan-Lindblad formalism, for the time (t) development of the density operator ρ(t) during thermalization, namely the process such that an open system with arbitrary initial states, when coupled to a thermal bath, ρ (t -> ∞) takes up a Gibbsian form. The theory is applied to a molecular vibrating system, to a semi-classical vibronic entity, and may be applied to the excitation probabilities in a condensed state’s phonon system and to arbitrarily large-scale systems (reaching as far as global warming). A sideline is to an entropy-decreasing Maxwell-demon type quantum state transition. While the prescription may not be unique, it gives rise to an experimentally testable non-monotonicity in the system’s information entropy. The calculated entropy maximum found for an electronic doublet is interpreted as a “transient democratization” of the states and, though lacking a formal proof, a conjecture is proposed for the occurrence of maxima in general instances.