Stochastic Information Research Articles

The value of pre-trip information on departure time and route choice in the morning commute under stochastic traffic conditions

Uncertainty in transportation systems can incur additional travel costs, but this adverse effect can be mitigated by providing travel information (Lindsey et al., 2014). It is not clear, however, if providing information always reduces system travel costs, particularly when the information provided is not one hundred percent accurate. This paper studies the welfare effects resulting from providing pre-trip information to morning commuters in a two-route network with bottlenecks, where the bottleneck capacity and free-flow travel time on each route are stochastic. We derive the expected travel costs at user equilibrium (UE) under stochastic conditions without and with pre-trip information first on one route, and later on two routes. Under this model, we examine how accurate and inaccurate information affect commuting costs. We find that full and accurate information is welfare-improving under stochastic bottleneck capacity and deterministic free-flow travel time. However, when bottleneck capacity and free-flow travel time are both stochastic, full and accurate information can be welfare-reducing for the two-route scenario. We show that the degree of correlation between routes in traffic conditions, the frequency and severity of bottleneck capacity drops, and the relationship between free-flow travel time and bottleneck capacity significantly affect the welfare effects of pre-trip information. Furthermore, when bottleneck capacity experiences severe drops, full information is more likely to be welfare-improving, even if the provided information is not completely accurate. These theoretical results are supplemented by numerical examples that show the welfare effects of providing pre-trip information to morning commuters.

Sequential Procurement with Contractual and Experimental Learning

We study the design of sequential procurement strategies that integrate stochastic and strategic information. We consider a buyer who repeatedly demands a certain good and is unable to commit to long-term contracts. In each time period, the buyer makes a price offer to a seller who has private, persistent information regarding his or her cost and quality of provision. If the offer is accepted, the seller provides the good with a stochastic quality that is not contractible. Therefore, the buyer can learn from the (strategic) acceptance decisions taken by the seller and from evaluations of the (stochastic) quality delivered whenever a purchase occurs. Hence, the buyer not only faces a tradeoff between exploration and exploitation but also needs to decide how to explore: by facilitating quality experimentation or by strategically separating seller types. We characterize the perfect Bayesian equilibria of this sequential interaction and show that the buyer’s equilibrium strategy consists of a dynamic sequence of thresholds on his or her belief on the seller’s type. When only one seller type is more efficient than the buyer’s outside option, the buyer uses one form of information: either strategic or stochastic. If both seller types are more efficient, then the buyer uses both forms of information; at the early stages of the interaction, the buyer offers high prices that incentivize trade and quality experimentation, and after gathering enough information, the buyer may advance to offering low prices that partially separate seller types. We identify the effect strategic sellers have on the buyer’s optimal strategy relative to more traditional learning dynamics and establish that, paradoxically, when sellers are strategic, the ability to observe delivered quality is not always beneficial for the buyer. This paper was accepted by Victor Martínez-de-Albéniz, operations management.

Hybrid market-based resources allocation in Mobile Edge Computing systems under stochastic information

In order to deal with the problem of user diversity in Mobile Edge Computing (MEC) resource trading market, in this paper, we propose a hybrid market-based resource transaction mechanism consisting of futures market and spot market. Two different types of users have been taken into consideration. One is registered users and another is unregistered users. In futures market, registered users pay a registration fee to the agent and use the reserved resources according to the contract signed exclusively. We design optimal contracts for the registered users by adjusting the registration fee in order to maximize the servers’ utility. In spot market, unregistered users compete with one another to purchase the resources on demand. We model the trading process as a multi-seller and multi-buyer market, and propose auction algorithms to match the asking price from servers and the bidding price from unregistered users by assigning computation resources to the users. The agent acts as the auctioneer to host the auction, and the unregistered users bid on computation resources based on the estimated valuation. We study the optimal solution under both complete and incomplete information scenarios, depending on whether the agent can observe the users’ private information. Simulation results demonstrate the existences of the asking price and registration fee for the servers to maximize utility.

Anticipatory approach for dynamic and stochastic shipment matching in hinterland synchromodal transportation

This paper investigates a dynamic and stochastic shipment matching problem faced by network operators in hinterland synchromodal transportation. We consider a platform that receives contractual and spot shipment requests from shippers, and receives multimodal services from carriers. The platform aims to provide optimal matches between shipment requests and multimodal services within a finite horizon under spot request uncertainty. Due to the capacity limitation of multimodal services, the matching decisions made for current requests will affect the ability to make good matches for future requests. To solve the problem, this paper proposes an anticipatory approach which consists of a rolling horizon framework that handles dynamic events, a sample average approximation method that addresses uncertainties, and a progressive hedging algorithm that generates solutions at each decision epoch. Compared with the greedy approach which is commonly used in practice, the anticipatory approach has total cost savings up to 8.18% under realistic instances. The experimental results highlight the benefits of incorporating stochastic information in dynamic decision making processes of the synchromodal matching system.

Chance-constrained stochastic MPC of Astlingen urban drainage benchmark network

In urban drainage systems (UDS), a proven method for reducing the combined sewer overflow (CSO) pollution is real-time control (RTC) based on model predictive control (MPC). MPC methodologies for RTC of UDSs in the literature rely on the computation of the optimal control strategies based on deterministic rain forecast. However, in reality, uncertainties exist in rainfall forecasts which affect severely accuracy of computing the optimal control strategies. Under this context, this work aims to focus on the uncertainty associated with the rainfall forecasting and its effects. One option is to use stochastic information about the rain events in the controller; in the case of using MPC methods, the class called stochastic MPC is available, including several approaches such as the chance-constrained MPC(CC-MPC) method. In this study, we apply CC-MPC to the UDS. Moreover, we also compare the operational behavior of both the classical MPC with perfect forecast and the CC-MPC based on different stochastic scenarios of the rain forecast. The application and comparison have been based on simulations using a SWMM model of the Astlingen urban drainage benchmark network. From the simulations, it was found that CSO volumes were larger when CC-MPC had overestimating forecast biases, while for MPC they increased with any presence of forecast biases.

Modelling regional-scale attenuation across Italy and the Tyrrhenian Sea

Modelling regional-scale attenuation of seismic waves at ~1 Hz is challenging, especially when these waves propagate across both continental and oceanic crust. Recent developments in seismic imaging and modelling have provided us with the computational tools necessary to reconstruct these mixed settings using deterministic (coherent) and stochastic (coda) information. Here, we present new tomographic maps of coda-attenuation for both the Italian peninsula and the Tyrrhenian Sea. Kernel-based coda attenuation imaging in the diffusive approximation is tested in the oceanic environment, highlighting a non-diffusive behaviour across the Southern Tyrrhenian Sea. Joint deterministic and Radiative Transfer forward modelling of coherent and scattered waves (Radiative3D) is then performed for the portion of this area showing the lowest coda attenuation. The corresponding parametric study shows that coda attenuation is an efficient marker of variations of Moho depths and crustal reverberations, such as those we observe in a transitional area (characterized by continental crust and thicker Moho) between two oceanic basins (Vavilov and Marsili). Diffusive tomographic imaging informed by Radiative3D modelling reconstructs high-attenuation patterns along the Italian peninsula in agreement with the thick sediments cover across the Po Plain and the Adriatic coast and the magmatic systems in the Central Campanian province. A combined approach of coda-attenuation imaging and modelling can constrain Moho depth and spatial variations of seismic attenuation at regional scale and in mixed continental-oceanic settings.

Stochastic resonance of volatility influenced by price periodic information in financial market

General researches show that all kinds of random risk information and periodic information in the financial system are mainly transmitted to the asset price through influencing the volatility, thus impacting the whole market. So can the periodic information and random factors in the price be transmitted to the volatility in reverse and cause volatility changes? Hence, in this paper, we investigate the stochastic resonance of volatility which is influenced by price periodic information in financial market, based on our proposed periodic Brownian Motion model and absolute return volatility. The parameter estimation of the periodic Brownian Motion model is obtained by minimizing the mean square deviation between the theoretical and empirical return distributions for the CSI300 data set. The good agreements of the probability density functions of the price returns, realized volatility (RV) at 5 minutes, RV at 15 minutes and absolute return volatility between theoretical and empirical calculation are found. After simulating the absolute return volatility and signal power amplification (SPA) of volatility via periodic Brownian Motion model, the results indicated that (i) single and double inverse resonance phenomena can be observed in the function of SPA versus random information intensity or economic growth rate; (ii) multiple inverse resonance phenomena can be also observed for SPA versus frequency of periodic information. The results imply that the transmission of stochastic factors and periodic information is not only from the volatility to the price, but also from the price to the volatility.

A Hybrid Stochastic Gradient Hamiltonian Monte Carlo Method

Recent theoretical analyses reveal that existing Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) methods need large mini-batches of samples (exponentially dependent on the dimension) to reduce the mean square error of gradient estimates and ensure non-asymptotic convergence guarantees when the target distribution has a nonconvex potential function. In this paper, we propose a novel SG-MCMC algorithm, called Hybrid Stochastic Gradient Hamiltonian Monte Carlo (HSG-HMC) method, which needs merely one sample per iteration and possesses a simple structure with only one hyperparameter. Such improvement leverages a hybrid stochastic gradient estimator that exploits historical stochastic gradient information to control the mean square error. Theoretical analyses show that our method obtains the best-known overall sample complexity to achieve epsilon-accuracy in terms of the 2-Wasserstein distance for sampling from distributions with nonconvex potential functions. Empirical studies on both simulated and real-world datasets demonstrate the advantage of our method.

Assessing the renewable energy efficiency levels of BRICS countries and Turkey using stochastic frontier analysis and information complexity criteria

Renewable energy is a sustainable energy source that can be produced repeatedly by using the resources that exist in nature's own evolution. Renewable energy sources occupy an important place in the world and our country due to their renewability, minimal environmental impact, low operating and maintenance costs, and their national qualifications, and reliable energy supply features. In this study renewable energy efficiency levels for the BRICS countries and Turkey were examined. In the study covering the period 2006-2015, we used the SFA method for efficiency analysis in input selection. We used information complexity criteria to decide which input set is the best on renewable energy efficiency process. The selection results pointed out to the CO 2 emission and Energy intensity as the most explanatory inputs. We observed that the selected inputs have significant effect on the renewable energy efficiencies. According to results, the renewable energy efficiency values follow approximately the same pattern for each country and do not vary significantly between the years. When comparing the renewable energy efficiencies among the countries, Brazil has the best performance with approximately 97% efficiency level, and Russia has the worst one. The efficiency level of Turkey is rather weak, but it is not the worst and the average efficiency is very close to China.

Information propagation in stochastic networks

In this paper, a network-based stochastic information propagation model is developed. The information flow is modeled by a probabilistic differential equation system. The numerical solution of these equations leads to the expected number of informed nodes as a function of time and reveals the relationship between the degrees of the nodes and their reception time. The validity of the model is justified by Monte Carlo network simulation through the analysis of information propagation in scale-free and Erdős–Rényi networks. It has been found that the developed model provides more accurate results compared to the widely used network-based SI mean-field model, especially in sparse networks.

A dual-randomness bi-level interval multi-objective programming model for regional water resources management

In this research, a dual-randomness bi-level interval multi-objective programming (DR-BIMP) model was developed for supporting water resources management among multiple water sectors under complexities and uncertainties. Techniques of bi-level multi-objective programming (BMOP), double-sided stochastic chance-constrained programming (DSCCP), and interval parameter programming (IPP) were incorporated into an integrated modeling framework to achieve comprehensive consideration of the complexities and uncertainties of water resources management systems. The DR-BIMP model can not only effectively deal with the interactive effects between multiple decision-makers in complex water management systems through the bi-level hierarchical strategies, but also can characterize the multiple uncertainties information expressed as interval format and probability density functions. It could thus improve upon the existing bi-level multi-objective programming through addressing discrete interval parameters and dual-randomness problems in optimization processes simultaneously. Then, the developed model was applied to a real-world case to optimally allocate water resources among three different water sectors in five sub-regions in the Dongjiang River basin, south China. The results of the model include determining values, interval values, and stochastic distribution information, which can assist bi-level decision-makers to plan future resources effectively to some extent. After comparing the variations of results, it is found that an increasing probability level can lead to higher system benefits, which is increased from [20,786.00, 26,425.92] × 108 CNY to [22,290.84, 27,492.57] × 108 CNY, while the Gini value is reduced from [0.365, 0.446] to [0.345, 0.405]. A set of increased probability levels gives rise to the lower-level objectives. Furthermore, the advantages of the DR-BIMP model were highlighted by comparing with the other models originated from the developed model. The comparison results indicated that the DR-BIMP model was a valuable tool for generating a range of decision alternatives and thus assists the bi-level decision-makers to identify the desired water resources allocation schemes under multiple scenarios.

Transformer-based identification of stochastic information cascades in social networks using text and image similarity

Identifying the origin of information posted on social media and how this may have changed over time can be very helpful to users in determining whether they trust it or not. This currently requires disproportionate effort for the average social media user, who instead has to rely on fact-checkers or other intermediaries to identify information provenance for them. We show that it is possible to disintermediate this process by providing an automated mechanism for determining the information cascade where a post belongs. We employ a transformer-based language model as well as pretrained ResNet50 model for image similarity, to decide whether two posts are sufficiently similar to belong to the same cascade. By using semantic similarity, as well as image in addition to text, we increase accuracy where there is no explicit diffusion of reshares. In a new dataset of 1,200 news items on Twitter, our approach is able to increase clustering performance above 7% and 4.5% for the validation and test sets respectively over the previous state of the art. Moreover, we employ a probabilistic subsampling mechanism, reducing significantly cascade creation time without affecting the performance of large-scale semantic text analysis and the quality of information cascade generation. We have implemented a prototype that offers this new functionality to the user and have deployed it in our own instance of social media platform Mastodon.

Stochastic Mutual Information Gradient Estimation for Dimensionality Reduction Networks.

Feature ranking and selection is a widely used approach in various applications of supervised dimensionality reduction in discriminative machine learning. Nevertheless there exists significant evidence on feature ranking and selection algorithms based on any criterion leading to potentially sub-optimal solutions for class separability. In that regard, we introduce emerging information theoretic feature transformation protocols as an end-to-end neural network training approach. We present a dimensionality reduction network (MMINet) training procedure based on the stochastic estimate of the mutual information gradient. The network projects high-dimensional features onto an output feature space where lower dimensional representations of features carry maximum mutual information with their associated class labels. Furthermore, we formulate the training objective to be estimated non-parametrically with no distributional assumptions. We experimentally evaluate our method with applications to high-dimensional biological data sets, and relate it to conventional feature selection algorithms to form a special case of our approach.

A deep learning based secured energy management framework within a smart island

This study proposes a novel secured management method for renewable microgrids considering the policies required for diagnosing cyber-attacks happening in the communication networks, usually applied in the secondary control layer of microgrids (MGs). Due to the so long stochastic and bad information entering the systems in order to make malicious attacks, their location and time data links have the ability of straying of those acting in normal conditions that attempt to have an effect on the precise voltage regulation and current dividing via influencing sensors of current and voltage. The ability to extract high-level features due to the usage of fast fourier transform (FFT) and deep learning (DL) for attack detection in cyberspace has made them to be considered as a strong technique in the face of small mutations or new attacks. These self-educated and compaction abilities of DL architectures have been considered as basic techniques for hidden scheme detection from the training datum for this reason attacks have been distinguished from benign traffic. A novel method, deep learning and FFT, for cyber-security has been used in the following paper with the aim of enabling the attacks detection in DC smart MG. The deep model and traditional machine learning way are evaluated in terms of performance, and distributed attack detection has been compared to the centralized diagnosing procedure. The tests proved that the distributed attack detection system studied can be more advanced in comparison with centralized detection systems applying FFT in the role of the input index of the DL model. This suggested distributed method enables for scalable monitoring of a MG and has the ability of detecting the existence of cyber-attacks in the communications between distributed generation agents (DGAs) controlled via a control on the basis of consensus and isolating the communication link over that the attack has been injected. Any local attack detection needs restricted information about its neighbor’s dynamics. The most important factor of the proposed detection plan can be that has the ability of detecting cyber-attacks with great precision and distinguishing cyber-attack from load changes.in addition, this has been shown that the suggested model can be further useful in the detection of the attack.

Intellect models used to search for data in stochastic information environment in Data Systems

IDEF models of the method of forming a semantic connection through calculations of the relevance on the basis of the area with the zone and attributes, and the calculation of the most cost-effective value of g, the method of calculating the weight of the data frequency and the frequency with inverse data, based on the model of intellectual semantic search, the creation of the ISIDR in the SIS and for software.*

Constrained Pythagorean Fuzzy Sets and Its Similarity Measure

Pythagorean fuzzy set (PFS) is an extension of the intuitionistic fuzzy set. It has a wider space of membership degrees. Thus, it is more capable in expressing and handling the fuzzy information in engineering practice and scientific research. However, PFSs lack a mathematical tool to express stochastic or probability information, rendering it unsuitable for application in many scenarios. In this article, an ordered number pair is used to describe fuzzy information and stochastic information under uncertain environments, namely constrained Pythagorean fuzzy set (CPFS). The CPFS has two components, <inline-formula><tex-math notation="LaTeX">$\text{CPFS}=(A,P)$</tex-math></inline-formula>, where <inline-formula><tex-math notation="LaTeX">$A$</tex-math></inline-formula> is the classical PFS, while <inline-formula><tex-math notation="LaTeX">$P$</tex-math></inline-formula> is a measurement of reliability for <inline-formula><tex-math notation="LaTeX">$A$</tex-math></inline-formula>. For PFS, CPFS is the first unified description of fuzzy information and probabilistic information, which is a more flexible way to describe knowledge or thinking. Furthermore, the similarity measure of CPFSs is presented, which meets the similarity measure theorems and can better indicate the flexibility of CPFSs. Numerical examples are used to demonstrate that the CPFSs similarity measure is reasonable and effective. The method of similarity measure can be degenerated to the similarity measure of PFSs under specific case and can avoid generating counter-intuitive results. In addition, similarity measure of CPFSs is applied to medical diagnosis and target classification of Iris. These experimental results have proven the practicability and effectiveness of our model.

Stochastic proximal splitting algorithm for composite minimization

Supported by the recent contributions in multiple domains, the first-order splitting became algorithms of choice for structured nonsmooth optimization. The large-scale noisy contexts make available stochastic information on the objective function and thus, the extension of proximal gradient schemes to stochastic oracles is heavily based on the tractability of the proximal operator corresponding to nonsmooth component, which has been highly exploited in the literature. However, some questions remained about the complexity of the composite models with proximal untractable terms. In this paper we tackle composite optimization problems, assuming only the access to stochastic information on both smooth and nonsmooth components, with a stochastic proximal first-order scheme with stochastic proximal updates. We provide sublinear \(\mathcal {O}\left( \frac{1}{k} \right) \) convergence rates (in expectation of squared distance to the optimal set) under the strong convexity assumption on the objective function. Also, linear convergence is achieved for convex feasibility problems. The empirical behavior is illustrated by numerical tests on parametric sparse representation models.

Formation of the concept of the socially responsible state as a factor of increasing the public governance and administration efficiency

The urgency of the topic is due to the fact that today the importance of socially responsible state as a factor in overcoming instability and crisis in COVID-19, which requires reformatting the principles of public administration, survival of organizations and their adaptation to changes in uncertainty in the environment and needs constant modernization. To form the concept of a socially responsible state in COVID-19, it is necessary to overcome the conditions of crisis and instability, to create conditions for sustainable development, reset the management of public administration, to form new survival values ​​of organizations at risk. The purpose of the study: theoretical and practical aspects of the concept the formation of socially responsible state as a factor in overcoming instability and crisis in COVID-19. Objectives of the study: 1. To study the formation of the concept of socially responsible state as a factor in overcoming instability and crisis in COVID-19. 2. To analyze the methodology of research of the socially responsible state as a factor of overcoming instability and crisis in the conditions of COVID-19. 3. To identify complexity as a sign of unstable and unbalanced public administration. 4. To identify the increasing complexity of organizations in modern conditions and the need for self-adaptation and self-regulation of systems. Research methodology. AGILE methodology acts as a system of new values, which extends to many areas of activity, including public administration and administration. Methods of synergetic analysis, including the theory of self-organizing open systems, which find themselves in points of imbalance, uncertainty, bifurcation (bifurcation), stochastic information (lack of information), system drift, resulting in a change of priorities. The result of the study. It is concluded that the state should realize and be responsible for its influence on society, promote socially responsible attitude to all processes taking place in society, for which to rebuild consciousness, principles and methods of activity that would be focused on obtaining social, economic and environmental effects that do not harm society and the environment.

Does the effort of Monte Carlo pay off? A case study on stochastic MPC

Abstract Stochastic Model Predictive Control (MPC) has established itself as a simple method to approximate stochastically disturbed optimal control problems, yet it comes with significant computational effort. In particular, taking into account all information of stochastic disturbances—if at all possible—can be expensive. However, due to the continuous re-optimization, MPC schemes also have a certain inherent robustness. Hence, we ask the question, which part of the perturbation is absorbed by the MPC-inherent robustness, and which needs to be explicitly taken into account in the optimization. In this paper, we compare several stochastic MPC algorithms taking into account a growing amount of the stochastic information: starting with deterministic schemes of certainty equivalence type, we use the noise’s distribution’s quantiles and finally we take, potentially, all stochastic information into account when performing Monte Carlo simulations in the MPC scheme. We carry out a numerical case study for the inverted pendulum, in which we observe that for short prediction horizons Monte Carlo does not pay off while for large prediction horizon it has slight advantages. Further, we discuss possible explanations for this behavior.

Route Duration Prediction in a Stochastic and Dynamic Vehicle Routing Problem with Short Delivery Deadlines

Abstract We are facing a real-world vehicle routing problem where orders arrive dynamically over the day at an online store and have to be delivered within short time. Stochastic information in form of the expected number and weight of orders and the traffic congestion level is available upfront. The goal is to predict the average time needed to deliver an order for a given time and day. This information is desirable for both routing decisions in the short horizon and planning vehicle drivers’ shifts with just the right capacity prior to the actual day. We compare a white box linear regression model and a neural network based black box model on historic route data collected over three months. We employ a hourly data aggregation approach with sampling statistics to estimate the ground truth and features. The weighted mean square error is used as loss function to favor samples with less uncertainty. A mean validation R2 score over 10 x 5-fold cross-validations of 0.53 indicates a substantial amount of unexplained variance. Both predictors are slightly optimistic and produce median standardized absolute residuals of about one.