A Penalized Sequential Convex Programming Approach for Continuous Network Design Problems

Mobile technologies and AI enable health data collection from devices, allowing effective monitoring. Traditional methods often compromise privacy, but federated learning (FL) offers a potential solution. However, current FL approaches face two issues: they don’t identify key health features for clinical intervention, and their aggregation overlooks patient multidimensional heterogeneity. This study seeks to develop a new FL method to tackle these challenges and enhance privacy in mobile health monitoring. This study proposes a novel FL method combining (1) a spatial and temporal attention-based prediction model (STA-Pred) that uses attention to identify key spatial and temporal features, and (2) a multidimensional heterogeneity-based aggregation protocol (MDH-Aggr), which aggregates components based on their heterogeneity to handle multidimensional differences. Experiments on three data sets show that our method outperforms existing methods in several patient-monitoring contexts. This study enhances understanding of how to leverage mobile technologies and AI to enable privacy-preserving health monitoring that promotes the social good. Additionally, it advances FL research through two innovative designs (STA-Pred and MDH-Aggr). History: This paper has been accepted by Kaushik Dutta for the Special Issue on Responsible AI and Data Science for Social Good. Funding: Y. Chai, H. Liu, and Y. Liu are supported by the National Natural Science Foundation of China [Grants 72342011, 72322019, 72188101, and 72402001]. Dr. L. Wang’s work was in part supported by the National Natural Science Foundation of China [Grant 72271027], Hainan Provincial Natural Science Foundation of China [Grant 726MS0458], and Beijing Institute of Technology Research Fund Program for Young Scholars [Grant XSQD-202216004]. X. Liu is not supported by any funds or associated with any of the abovementioned funds. Supplemental Materials: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.0521 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2023.0521 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/.

This paper develops a cost-effective sequential route recommender system to provide real-time routing recommendations for vacant taxis searching for the next passenger. We propose a prediction-and-optimization framework to recommend the searching route that maximizes the expected profit of the next successful passenger pickup based on the dynamic taxi demand-supply distribution. Specifically, this system features a deep learning-based predictor that dynamically predicts the passenger pickup probability on a road segment and a recursive searching algorithm that recommends the optimal searching route. The predictor integrates a graph convolution network (GCN) to capture the spatial distribution and a long short-term memory (LSTM) to capture the temporal dynamics of taxi demand and supply. The GCN-LSTM model can accurately predict the pickup probability on a road segment with the consideration of potential taxi oversupply. Then, the dynamic distribution of pickup probability is fed into the route optimization algorithm to recommend the optimal searching routes sequentially as route inquiries emerge in the system. The recursion tree-based route optimization algorithm can significantly reduce the computational time and provide the optimal routes within seconds for real-time implementation. Finally, extensive experiments using Beijing Taxi GPS data demonstrate the effectiveness and efficiency of the proposed recommender system. History: Accepted by Ram Ramesh, Area Editor for Data Science and Machine Learning. Funding: This work was partially supported by the Hong Kong Research Grants Council [Grants CityU 21500220, CityU 11504322] and the National Natural Science Foundation of China [Grant 72201222]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2021.0112 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2021.0112 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Infodemic is a significant concern for society, and numerous studies have explored artificial intelligence (AI)-based solutions to manage it. However, existing methods fail to make sense of the complex situation of infodemic and fail to manage infodemic in a proactive manner. To tackle this challenge, we formalize infodemic topic prediction (ITP) as a distinct research problem. ITP aims to identify emerging topics to provide a global view of the situation, as well as to predict the probability of a topic becoming an infodemic to enable proactive management. Although some relevant methods exist, they are insufficient in modeling the intertopic relationships where each topic is closely connected to all others in a group of topics (called multilateral intertopic relationships). This study proposes a new hypergraph-based ITP framework that models the multilateral intertopic relationship with a novel hypergraph. Specifically, we introduce a novel temporal topic hypergraph (TTH) where topics are treated as nodes, and the multilateral intertopic relationships in both semantic and temporal perspectives are modeled by hyperedges. The main novelties are twofold. First, our TTH is a novel hypergraph that can cope with newly emerged topics through the proposed directed and undirected hyperedges. Second, we propose a novel similarity-based transformation method (STM) that reduces the complexity of hypergraph transformation from [Formula: see text] to [Formula: see text], making it scalable for social media data. Evaluations on infodemics during the COVID-19 and the Mpox pandemic demonstrate the effectiveness of our framework. This study contributes to responsible infodemic management by formalizing the task of ITP and introducing a novel framework for it, which enables public health organizations, social media platforms, and policymakers to proactively make sense of and effectively respond to infodemic. History: This paper has been accepted by Kaushik Dutta for the Special Issue on the Responsible AI and Data Science for Social Good. Funding: L. Zhao, S. Ding, Y. Chai and S. Yang are supported by the National Natural Science Foundation of China [Grants 72293581, 72188101, 72342011, and 72322019]. J. Xie and X. Fang are not supported by any funds and are not associated with any of the above mentioned funds. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.0660 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.0660 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Ranking problems are commonly encountered in practical applications, including order priority ranking, wine quality ranking, and piston slap noise performance ranking. The responses of these ranking applications are often considered as continuous responses, and there is uncertainty on which scoring function is used to model the responses. In this paper, we address the scoring function uncertainty of continuous response ranking problems by proposing a ranking model averaging (RMA) method. With a set of candidate models varied by scoring functions, RMA assigns weights for each model determined by a K-fold crossvalidation criterion based on pairwise loss. We provide two main theoretical properties for RMA. First, we prove that the averaging ranking predictions of RMA are asymptotically optimal in achieving the lowest possible ranking risk. Second, we provide a bound on the difference between the empirical RMA weights and theoretical optimal ones, and we show that RMA weights are consistent. Simulation results validate RMA superiority over competing methods in reducing ranking risk. Moreover, when applied to empirical examples—order priority, wine quality, and piston slap noise—RMA shows its effectiveness in building accurate ranking systems. History: Accepted by Ram Ramesh, Area Editor for Data Science & Machine Learning. Funding: This research was supported by the Beijing Municipal Natural Science Foundation [Grant 1222002], the National Natural Science Foundation of China [Grants 12071457, 12201018, 12301364, 71925007, 72091212, and 72273120], National Quality Infrastructure [Grant 2022YFF0609903], the Chinese Academy of Sciences Project for Young Scientists in Basic Research [Grant YSBR-008], and the Natural Science Foundation of Anhui Province [Grant 2308085QA09]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.0257 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2023.0257 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

The redundancy allocation problem (RAP) aims to find an allocation of redundant components that maximizes system reliability under resource constraints. In this paper, we consider the RAP for complex systems with heterogeneous components, which may have bridges or interconnecting subsystems and are generally more complicated in structure than commonly studied series-parallel systems. A major challenge of the RAP is that the reliability function is nonlinear. To tackle this, we develop six exact mixed integer linear programming (MILP) formulations, which, to our knowledge, are the first of such formulations for the problem. The key idea is to use binary variables and big-M constraints to successively linearize each product term in the reliability function. In addition, we propose a novel tree representation of the reliability function to establish the interconnections between different product terms, which is then used to reduce the number of decision variables and constraints needed in the linearization. We also propose an alternative form of the reliability function that only involves the failure probabilities of subsystems, and leverage this to derive better variable bounds, smaller big-M constants, and simpler formulations. We compare the six MILP formulations by solving them directly using an MILP solver based on three sets of test instances of varying complexity. Results indicate that all formulations achieve optimal solutions quickly for the first set and within reasonable computation times for the second set. For the third set, two formulations can always be solved within a given time limit. Overall, the MILP formulation derived based on the alternative form of the reliability function and a specific type of tree representation scheme performs the best. We also demonstrate that our approach is more robust and scalable than a benchmark branch-and-bound method. History: Accepted by Pascal Van Hentenryck, Area Editor for Computational Modeling: Methods & Analysis. Funding: This work was supported by National Natural Science Foundation of China [Grants 72310107003, 72471189, 71901176]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.0842 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.0842 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

In this paper, we present an efficient scenario reduction method for optimization problems whose objective function is the higher moment coherent risk (HMCR) measures. Compared with existing approaches, our method places greater emphasis on the characteristics of optimization problems. Because the value of HMCR measures only depends on a small subset of scenarios that correspond to high cost or loss, our approach is based mostly on the concept of ineffective scenarios previously proposed in the literature, which entails identifying the scenarios whose removal from the problem results in no change of the optimal value. We test our method on a simple portfolio optimization problem with only nine risky assets and a realistic one with 50 risky assets and cardinality constraints. Results show that our scenario reduction method can yield a more accurate optimal solution and optimal value, along with a smaller reduced-scenario set. Even at the same reduction level, our method continues to outperform the existing scenario reduction methods. Interestingly, the portfolio produced by our method is also more diversified than others. History: Accepted by Pascal Van Hentenryck, Area Editor for Computational Modeling: Methods & Analysis. Funding: This work was supported by the National Natural Science Foundation of China [Grants 71720107002 and U1901223]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.0375 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2022.0375 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Sparse reduced-rank regression is a widespread tool to reveal the association between multiple responses and predictors, and it has been widely applied to many data-driven applications. Although much of the literature has studied related theoretical properties and numerical algorithms, due to high nonconvexity, the computation burden for large-scale data sets remains a great challenge in practice. Also, the gap between the statistical consistency and the algorithmic convergence needs more research. To address these two issues, we formulate a sparse reduced-rank regression as a set of parallel cosparse unit-rank estimation problems and propose a new algorithm to estimate these subproblems in parallel. Under mild conditions, the iteration complexity of the proposed algorithm is polynomial with high-dimensional responses and predictors. We show a statistical consistency for the numerical solution, thereby bridging the gap between statistical consistency and numerical computation from nonconvex optimization. Moreover, the main calculation of the algorithm is restricted to a small active set, so it exhibits fast computation even in high dimensions. Extensive numerical studies and an application in genetics demonstrate the effectiveness and scalability of our approach. History: Accepted by Antonio Frangioni, Area Editor for Design & Analysis of Algorithms–Continuous. Funding: This work was supported by the National Key R&D Program of China [Grant 2024YFA1012200], the National Natural Science Foundation of China [Grants 12171449 and 72401266], the Fundamental Research Funds for the Central Universities [Grant WK2040000079], the USTC Research Funds of the Double First-Class Initiative [Grant YD2040002019], and the China Postdoctoral Science Foundation [Grant 2023M733402]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.0691 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.0691 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

We present a new difference of convex functions algorithm (DCA) for solving linear and nonlinear mixed complementarity problems (MCPs). The approach is based on the reformulation of bilinear complementarity constraints as difference of convex (DC) functions, more specifically, the difference of scalar, convex quadratic terms. This reformulation gives rise to a DC program, which is solved via sequential linear approximations of the concave term using standard DCA techniques. The reformulation is based on a generalization of earlier results on recasting bilinearities and leads to a novel algorithmic framework for MCPs. Through extensive numerical experimentation, the proposed approach, referred to as DCA-BL, proves to be an efficient heuristic for complementarity problems. For linear complementarity problems (LCPs), we test the approach on a number of randomly generated instances by varying the size, the density, and the eigenvalue distribution of the LCP matrix, providing insights into the numerical properties of DCA-BL. In addition, we apply the framework to a market equilibrium problem and find that DCA-BL scales well on realistic LCP instances. Also, through experimentation, we find that that DCA-BL performs particularly well compared with other DC-based complementarity approaches in the literature if the LCP is highly dense, asymmetric, or indefinite. Lastly, the method is successfully applied to a set of equilibrium problems with second-order cone constraints, which give rise to nonlinear complementarity problems, with applications to stochastic equilibrium problems in water infrastructure and finance. History: Accepted by Antonio Frangioni, Area Editor for Design & Analysis of Algorithms–Continuous. Funding: This work was supported by the Division of Civil, Mechanical and Manufacturing Innovation [Grant 2113891], the Division of Electrical, Communications and Cyber Systems [Grant 2114100], the Office of Naval Research Global [Grant 00014-22-1-2649], Petrobras [Grant 4324713], the Division of Mathematical Sciences [Grant 2318519], and the Danmarks Frie Forskningsfond [Grant 0217-00009B]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2024.0822 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2024.0822 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

Defined solely by means of order-theoretic operations meet (min) and join (max), weighted lattice polynomial functions are particularly useful for modelling data on an ordinal scale. A special case, the discrete Sugeno integral, defined with respect to a nonadditive measure (a capacity), enables accounting for the interdependencies between input variables. However, until recently the problem of identifying the fuzzy measure values with respect to various objectives and requirements has not received a great deal of attention. By expressing the learning problem as the difference of convex functions, we are able to apply DC (difference of convex) optimization methods. Here we formulate one of the global optimization steps as a local linear programming problem and investigate the improvement under different conditions. Abbreviation: DC: difference of convex; LP: linear programming

In this paper, we consider a method to solve the device-free localization (DFL) problem that is able to detect spatial obstruction via wireless network. A dictionary learning approach with difference of convex (DC) programming and DC algorithm is proposed to indicate target location based on learning data. By measuring the variation in the received signal strength of the wireless links indicating the locations of the obstructions, the physical target in the monitoring area can be estimated through a learned dictionary. We show that the DFL problem can be efficiently formulated as a non-convex optimization problem. We adopt a penalty function called the minimax concave penalty, which possesses good properties in terms of seeking sparsity, and solve the non-convex optimization problem using DC programming. Furthermore, the localization accuracy achieved during the path-tracking task is further improved by the proposed tracking neighborhood rule. The rule provides a solution for increasing the localization accuracy under time-varying conditions generated by sampling channels of sensor networks under noisy conditions. The proposed approach is validated on a real-world dataset and has the potential to be adopted flexibly in DFL applications.

We propose a preference-learning algorithm for uncovering Decision Makers’ (DMs’) contingent evaluation strategies in the context of multiple criteria sorting. We assume the preference information in the form of holistic assignment examples derived from the analysis of alternatives’ performance vectors and textual descriptions. We characterize the decision policies using a mixture of threshold-based, value-driven preference models and associated latent topics. The latter serve as the stimuli underlying the contingency in decision behavior. Such a probabilistic model is constructed by using a flexible and nonparametric Bayesian framework. The proposed method adopts a hierarchical Dirichlet process as the prior so that a group of DMs can share a countably infinite number of contingent models and topics. For all DMs, it automatically identifies the components representing their evaluation strategies adequately. The posterior is summarized by using the Hamiltonian Monte Carlo sampling method. We demonstrate the method’s practical usefulness in a real-world recruitment problem considered by a Chinese IT company. We also compare the approach with counterparts that use a single preference model, implement the parametric framework, or consider each DM’s preferences individually. The results indicate that our approach performs favorably in both interpreting DMs’ contingent decision behavior and recommending decisions on new alternatives. Furthermore, the approach’s performance and robustness are investigated through a computational experiment involving real-world data sets. History: Accepted by Ram Ramesh, Area Editor for Data Science and Machine Learning. Funding: J. Liu received financial support from the National Natural Science Foundation of China [Grants 72071155 and 71701160]. M. Kadziński received financial support from the Polish National Science Center under the SONATA BIS project [Grant DEC-2019/34/E/HS4/00045]. X. Liao received financial support from the National Natural Science Foundation of China [Grant 71872144]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.1292 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2021.0328 ) at ( http://dx.doi.org/10.5281/zenodo.7608750 ).

An intelligent resource management scheme for SDF-based cooperative spectrum sensing in the presence of primary user emulation attack

The security requirements of data transmission over wireless networks are energy-limited in many situations. In this paper, the secure energy efficiency (EE), is defined as the ratio of the secrecy rate to the total power, and is investigated in a systematic way considering a decode-and-forward (DF) relay network with a potential eavesdropper. We maximize the secure EE subject to the individual power constraint and the minimum decoding rate constraint of the relay. To deal with the nonconvexity of the formulated problem, a fractional programming approach embedded with DC (difference of convex functions) programming is proposed to solve the problem by two-layer iterations. The key point of the proposed algorithm is to translate the primal problem into a series of convex subproblems, which can be solved by convex programming. It is verified by simulation that the proposed algorithm achieves much better secure EE than the conventional secrecy rate maximization yet with a minor performance loss measured by average secrecy rate or secrecy outage probability.

We consider the optimal sourcing problem when the available suppliers are subject to ambiguously correlated supply risks. This problem is motivated by the increasing severity of supply risks and difficulty evaluating common sources of vulnerability in upstream supply chains, which are problems reported by many surveys of goods-producing firms. We propose a distributionally robust model that accommodates (i) multiple levels of supply disruption, not just full delivery or no delivery, and (ii) can use data-driven estimates of the underlying correlation to develop sourcing strategies in situations where the true correlation structure is ambiguous. Using this framework, we provide analytical results regarding the form of a worst-case supply distribution and show that taking such a worst-case perspective is appealing due to severe consequences associated with supply chain risks. Moreover, we show how our distributionally robust model may be used to offer guidance to firms considering whether to exert additional effort in attempt to better understanding the prevailing correlation structure. Extensive computational experiments further demonstrate the performance of our distributionally robust approach and show how supplier characteristics and the type of supply uncertainty affect the optimal sourcing decision. History: Accepted by Pascal Van Hentenryck, Area Editor for Computational Modeling: Methods and Analysis. Funding: K. Pan was supported in part by the Research Grants Council of Hong Kong [Grant 15501920] and in part by the National Natural Science Foundation of China [Grant 72001185]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2022.1254 ) or is available from the IJOC GitHub software repository ( https://github.com/INFORMSJoC ) at http://dx.doi.org/10.5281/zenodo.7129356 .

Semi-supervised learning is especially important in data mining applications because it can make use of plentiful unlabeled data to train the high-quality learning models. Semi-Supervised Support Vector Machine (S3VM) is a powerful semi-supervised learning model. However, the high computational cost and non-convexity severely impede the S3VM method in large-scale applications. Although several learning algorithms were proposed for S3VM, scaling up S3VM is still an open problem. To address this challenging problem, in this paper, we propose a new incremental learning algorithm to scale up S3VM (IL-S3VM) based on the path following technique in the framework of Difference of Convex (DC) programming. The traditional DC programming based algorithms need multiple outer loops and are not suitable for incremental learning, and traditional path following algorithms are limited to convex problems. Our new IL-S3VM algorithm based on the path-following technique can directly update the solution of S3VM to converge to a local minimum within one outer loop so that the efficient incremental learning can be achieved. More importantly, we provide the finite convergence analysis for our new algorithm. To the best of our knowledge, our new IL-S3VM algorithm is the first efficient path following algorithm for a non-convex problem (i.e., S3VM) with local minimum convergence guarantee. Experimental results on a variety of benchmark datasets not only confirm the finite convergence of IL-S3VM, but also show a huge reduction of computational time compared with existing batch and incremental learning algorithms, while retaining the similar generalization performance.