Search

This paper provides a general representation of endogenous and threshold-based regime switching models and develops an efficient numerical solution method. The regime-switching is triggered endogenously when some variables cross threshold conditions that can themselves be regime-dependent. We illustrate our approach using a RBC model with state-dependent government spending policies. It is shown that regime-switching models involve strong non linearities and discontinuities in the dynamics of the model. However, our numerical solution based on simulation and projection methods with regime-dependent policy rules is accurate, and fast enough, to efficiently take into all these challenging aspects. Several alternative specifications to the model and the method are studied.

The aim of this paper is two-fold. The first part is to generalize the simple regret model by Bell (1982) and Loomes and Sugden (1982) to cope with the situation in which decision outcomes are multi-attributed. We propose a model which combines the simple regret model for ex ante preferences and the additive difference representation for ex post preferences. Then we show a necessary and sufficient axiomatization of our model under Savage’s framework. The proposed model is composed of three types of functions. One is a value function for each attribute. The others are attribute-dependent and holistic regret-rejoicing functions which respectively capture trade-offs among value-differences of chosen and forgone outcomes for each attribute and among all attributes. We also discuss constructive procedures and their consistency. The second part is to conduct an experiment to measure two attributes case. Participants were asked to join in a work to carry over heavy bags at some varying distances. The data supports the framing effect, that is, participants tend to have convex value functions in each attribute and indicate regret averse for weights and regret seeking for distances in gain frame. On the other hand, participants exhibit concave value functions for both attributes and regret seeking for weights and regret averse (almost regret neutral) for distances in loss frame.

A fast and accurate representation of the complex heat and mass transfer processes of closed-wet cooling tower (CWCT) systems for optimal design and management is challenging, especially considering the impact of weather variability. This study develops a systematic approach for the stochastic bi-objective optimization of CWCT systems based on a simplified analytical model. The analytical model is constructed by using the effectiveness-number of transfer unit method, which is embedded in a stochastic non-linear programming model for optimizing two competing objectives, namely the specific area and the coefficient of performance. A decomposition strategy is proposed to convert the original problem into multiple bi-objective optimization sub-problems by sequentially discretizing the design and uncertain spaces. Each bi-objective optimization sub-problem is solved using an augmented e-constraint method. The resulting solutions are finally processed by an efficient multi-criteria decision-making method, namely the Technique for Order Preference by Similarity to Ideal Solution to select the preferred Pareto solution. From an illustrative case study, it is found that the stochastic approach outperforms the deterministic approach since it allows for a certain design margin of heat exchanger area to reduce the impact of the variability in weather and to avoid the risks of abnormal running cases.

The classical way to get an analytical model for the (supposedly heavy) tail of a loss severity distribution is via parameter inference from empirical large losses. However, in the insurance practice it occurs that one has much less information, but nevertheless needs such a model, say for reinsurance pricing or capital modeling. We use the Generalized Pareto distribution to build consistent underlying models from very scarce data like: the frequencies at three thresholds, the risk premiums of three layers, or a mixture of both. It turns out that for typical real-world data situations such GPD “fits” exist and are unique. We also provide a scheme enabling practitioners to construct reasonable models in situations where one has even less, or somewhat more, than three such bits of information. Finally, we have a look at model risk, by applying some parameter-free inequalities for distribution tails and a particular representation for loss count distributions. It turns out that, in the data situation given above, the uncertainty about the severity can be surprisingly low, such that the overall uncertainty is driven by the loss count.

Traditionally, the discovery of new solid-state materials with user-defined properties is driven either by human intuition, heuristic chemical rules, and/or density functional theory (DFT). However, these methods have limitations, either in accessibly (domain expertise), accuracy, and/or throughput [1]. Consequently, the material search space remains underexplored, given order 105 reported compounds compared to order 1010 theorized ternary compounds [2]. To accelerate the exploration of new solid-state materials, a framework capable of inverse design for materials with user-defined properties is needed. Herein, we present a generalized framework for inverse design of crystals with user-defined properties, which include both ground-state and excited-state properties (e.g., thermoelectric power factor) using sparsely labelled training data. The key enabler of this inverse-design framework is a general and invertible crystallographic representation that encodes the crystallographic information into the representations in both real space and reciprocal space. The trained surrogate model achieves similar property prediction accuracy and precision as DFT calculations within seconds. Using the developed inverse-design framework, we design 79 new crystals with user-targeted formation energies, 17 crystals with targeted bandgap, and 27 crystals for potential thermoelectric applications. The compositions of those designed materials are unique and cannot be found in the training or test sets. We validate our predictions using first-principle calculations. ­­Toward bridging the gap between simulation and experiment, we demonstrate a naive synthesizability metric — predicting the existence of an ICSD record — and show this methodology can, in principle, include stability and/or synthesizability as a target metric, once consensus metrics are agreed upon by the field.

The hybrid analytics and modelling tool is one of the key techniques for the digital twin of industrial processes. Attempt for hybridization of the thermochemical and thermophysical properties of molten SiO2-Al2O3-CaO slags has been recently made at SINTEF Industry. The power of the deep neural networks for the multiphysical modelling of ferroalloy production will be demonstrated in the paper. A deep neural network has been set up for representing thermophysical (electronic conductivity, liquidus temperature, viscosity as well as density) and thermochemical (slag-metal equilibrium composition) properties of the SiO2-Al2O3-CaO system. The physics-based models were first used to model the density, conductivity and viscosity of molten SiO2-Al2O3-CaO slags. Liquidus temperature and metal-slag equilibrium data were calculated using the well-known thermochemical software package, FactSage. These physical models can be used as high-throughput calculation tools for the generation of industrial important labelled data. The DNN model can represent the above labelled metadata with a set of optimized parameters. The DNN model developed in the present study consists of 4 input variables (composition and temperature), 1 hidden layer and 5 output properties was used. The DNN model calculation can almost entirely reproduce all above properties using about 120 parameters. Furthermore, the DNN model can be automatically updated by continuously feeding the experimental data from the laboratory and industry (machine learning). This provides a sound basis for the development of the localized HAM tool for individual ferroalloy producers.

Transfer learning, the application of knowledge from a training domain to a novel domain, is a critical advantage for human intelligence and an eagerly sought goal for artificial intelligence. Here, we describe a neural mechanism for transfer learning based on compositional representation. In monkeys trained on a small alphabet of letter-like stimuli, neurons in high-level visual cortex showed increased selectivity specifically for compositional letter elements (limbs, junctions, terminations) and their interactions (combinations of limbs, etc.) and not for whole letters. Strikingly, the enhanced selectivity for the familiar compositional elements and their interactions transferred to novel letters constructed from the same elements, making the novel letters more discriminable and providing information about their structure. Compositionality is a powerful strategy for knowledge transfer in domains like vision where objects often share common elements. Our results reveal a brain mechanism for visual transfer learning that could be adapted for artificial vision.

SummaryAs the only nonlinear and most diverse biological sequence, glycans offer substantial challenges for computational biology. These complex carbohydrates participate in nearly all biological processes – from protein folding to the cellular entry of viruses – yet are still not well understood. There are few computational methods to link glycan sequences to functions and those that do exist do not take full advantage of all the available information of glycans. SweetNet is a graph convolutional neural network model that uses graph representation learning to facilitate a computational understanding of glycobiology. SweetNet explicitly incorporates the nonlinear nature of glycans and establishes a framework to map any glycan sequence to a representation. We show that SweetNet outperforms other computational methods in predicting glycan properties on all reported tasks. More importantly, we show that glycan representations, learned by SweetNet, are predictive of organismal phenotypic and environmental properties. Finally, we present a new application for glycan-focused machine learning, the prediction of viral glycan-binding, that can be used to discover new viral receptors and monitor rapidly mutating viruses.

The bucketed PCA neural network (PCA-NN) with transforms is developed here in an effort to benchmark deep neural networks (DNN's), for problems on supervised classification. Most classical PCA models apply PCA to the entire training data set to establish a reductive representation and then employ non-network tools such as high-order polynomial classifiers. In contrast, the bucketed PCA-NN applies PCA to individual buckets which are constructed in two consecutive phases, as well as retains a genuine architecture of a neural network. This facilitates a fair apple-to-apple comparison to DNN's, esp. to reveal that a major chunk of accuracy achieved by many impressive DNN's could possibly be explained by the bucketed PCA-NN (e.g., 96% out of 98% for the MNIST data set as an example). Compared with most DNN's, the three building blocks of the bucketed PCA-NN are easier to comprehend conceptually - PCA, transforms, and bucketing for error correction. Furthermore, unlike the somewhat quasi-random neurons ubiquitously observed in DNN's, the PCA neurons resemble or mirror the input signals and are more straightforward to decipher as a result.

One of the prerequisites in all electrochemical processes is to achieve uniform distribution in potential and current acoss the electrodes and the cell. Although parallel plate electrodes encompass a preferable uniformity in distribution, it tends to get attenuated based on the geometric variables like electrode spacing and width of electrrodes. The purpose of this paper is to analyse underlying aspects associated with these distributions and to indicate how they can be visually represented. A mathematic treatment is carried out to generate an expression for electrolyte potential at the electrode surface with varying ratio of width of electrode to electrode spacing (L/H). Segregation of L/H, where improved uniformity could be attained, is evaluated based on a graphical study. In the latter part, simulation studies were conducted to examine the profile of both current and potential in a typical parallel plate electrochemical cell so as to get a visual representation. This study is proposed to be a cornerstone for practicing electrochemical engineers in designing electrodes and electrochemical cells for various applications.

It has long been observed that model structure may play an important role in determining the effectiveness of certain solution techniques to mixed integer linear programs (MILPs). One implication is that solution techniques that work well for one class of MILP are often found to be effective on other “similar” problems. Despite the potential benefits, there is limited research in the area of recognizing similarity among MILPs. This paper seeks to address this gap by presenting a framework for classifying and comparing instances of MILPs, based on their mathematical structure. A new graph-based representation of MILPs is proposed, where decision variables and constraints are described by nodes, and arcs denote the presence of decision variables in certain constraints. Inspired by recent advances in the area of deep learning for graph-structured data, two methods of leveraging the MILP graph representations for classification and comparison are provided. In the first relational approach, a graph convolutional network (GCN) is employed to classify these so-called MILP graphs as having come from one of a known number of problem types. The second approach makes use of latent features learned by the GCN to compare graphs to one another directly. As part of the latter approach, we introduce a formal measure of graph-based structural similarity. Empirical studies indicate that both the classification and comparison procedures perform well on a test set. Additional properties of MILP graphs are also explored through a series of case studies.

We start by showing that it it is possible to build a general class of interpolators which preserve local constantness and are both strictly local and infinitely smooth. We then leverage that class, called core, to produce a hierarchical representation of the signal by adding two differential layers to it. The second (resp. first) layer is designed in such a way that it uses only second (resp. first) order derivative samples and insures a perfect reconstruction of polynomial signals of degree up to two (resp. one).

This paper provides a new, unified, and flexible framework to measure and characterize a convergence process. Specifically, we formally define a general notion of asymptotic price convergence, which encompasses convergence as steady-state and catching-up, and propose a model to represent a wide range of transition paths. Our framework enables the econometric measurement of such transitional behaviors and the development of testing procedures. In particular, we use unit root tests (with and without breaks) to determine whether asymptotic convergence exists and, if so, its starting point and its type: as steady-state or catching-up and in its weak or strong version. We apply this methodology to illustrate the convergence process in two very different cases: (1) the aggregated price level of the Eurozone’s largest economies (Germany, France and Italy) and, (2) the transatlantic 19th Century wheat prices.

The advent of quantitative investing has made it increasingly important to understand the performance drivers of systematic strategies that use derivatives, such as those based on the sale of options. In this paper we introduce a new formulaic representation to analyse the performance of delta-hedged vanilla options and use this formula to break down the historical performance of foreign exchange straddles held to maturity. We show how, as is widely assumed, the performance is dictated in part by the so-called volatility premium, but also that another important driver is at play, the covariance between the convexity of the option and the prevailing volatility regime. In particular, we determine empirically that the sign of that driver’s contribution is tenor specific. These results can be used to size directional or relative value option strategies and to disentangle the impact of the various performance drivers in the booming field of systematic options strategies.

We introduce range utility theory for decisions under risk. Two functions are implicated in the representation of preferences: a traditional utility function for wealth---or changes in wealth---and a range distortion function. The latter introduces a local deformation of the utility function on the range under consideration, possibly magnifying the sensitivity to intermediate outcomes. In a sense we make precise, range utility theory is an alternative way to extend the binary rank dependent model to more than two outcomes. It can account for the fourfold pattern, as well as for the preference reversal phenomenon. When the range is fixed, range utility is linear in probabilities. We illustrate this advantage in game theory applications.

One goal of the project described in this paper is to create learning algorithms for machines and robots that lack a precise virtual controller for correct simulations. Using a digital twin approach, the developed mixed reality application aims for an overlay of a virtual robot model with the real world counterpart using Microsoft HoloLens 2 smart glasses. The application should help users to have an inside look into the results of the learning algorithm and therefore supervise and improve those results. The main focus of this paper is the visual representation of the digital twin on the smart glasses. One of the challenges is the level of abstraction and specific use of shaders (program code defining material attributes) to help the user differentiating between virtual and real objects. Therefore different presentation methods are described and evaluated. Study results with 48 persons show that the most abstract representation (wireframe) scores lowest, whereas a half-transparent model works best.

We prove suboptimality of probability matching in prediction tasks with an arbitrary (finite) number of outcomes and repetitions. For the popular case of binary prediction tasks, we also provide a graphical representation of the result. Finally, we relate probability matching to impulse balance equilibrium theory and show when probability matching is consistent with its predictions.

Data knowledge envelops the discovery of information. This includes the usage of measurements and data thoughts. Explanatory statistics can be used to summarize the data whereas inferential statistics can be utilized to survey two data sets. Regressions can be exploited to discover the connections between various variables. Data visualizations use charts, graphs, and dashboards to understand the data. This phase allows you to understand the quality of data. The client would then be able to choose information scopes of the chose factors and start a bunch examination. The bunched information are then pictured as a lot of circles. In an extra perception, we first sort information estimations of a customer variable and afterward picture the arranged cubic information in a shape utilizing volume rendering and iso surfaces. The intuitive connection representation permits showcasing analysts to rapidly investigate a wide range of blends of factors. This empowers them to discover important customer personal conduct standards a lot quicker. The grouped perception permitted analysts to distinguish itemized gatherings of worker with comparable conduct. Also, the representation of the arranged cubic information gives in depth data of one variable over the all. With these representations, a superior comprehension is given on worker conduct. R is utilized for investigating huge information in distributed computing. The point is to recognize the difficulties for breaking down huge information. R is a factual programming language which is behind measurements, examination and perception. The present information researcher and business pioneer utilizes R to settle on power business choices.

Downside and deviation risk measures are becoming more and more important in many disciplines with clear interfaces with Applied Mathematics and Operations Research. Their dual representations have played critical roles in most of their applications (risk management, portfolio selection, pricing and hedging, etc.), but, to the best of our knowledge, bidual representations were never profoundly studied. New linear bidual representations will be provided, and their great capacity to linearize many problems will be proved, with special focus on risk optimization. This is important because there are very tractable necessary and sufficient optimality conditions and resolution algorithms in Linear Programming. Moreover, in the linearization process, one will have to introduce new decision variables providing us with very important information, such as sensitivities with respect to the selected risk measure and sensitivities with respect to the selected model (model risk). The theory will be presented for general Banach spaces, and an illustrative example will be given.

Machine learning and especially deep learning architectures provide a fresh perspective on the study of many body physics phenomena. In this paper, we employ Restricted Boltzmann machines (RBM) to represent quantum many-body states and find connections that can be made useful to quantum many-body physics research, ultimately leading to a better understanding of the fundamental nature of entanglement entropy in quantum physics. In this work, we establish the conditions for translating RBMs into Matrix Product States (MPS), showing that deep learning algorithms can be exploited as a powerful tool for an efficient representation of quantum states. We present an algorithm for mapping an RBM into an MPS, with a specific proof for Ising model. We discuss the upper entropy bound and entanglement properties resulting from such a connection, together with the consequences of our results in a broader context.

Biometrics is a technology that is currently being used in diverse aspects of life for hassle free authentication and verification. Biometric finds itself in almost every critical infrastructure secure access systems thus stepping on as a surrogate representation of an individual. The flip side to this surrogate representation is that if its compromised, it implies an identity theft of an individual. To circumvent this issue of identity theft in general and biometric template security in particular, we propose and implement a new fused biometric iris encryption algorithm for the generation of a secure biometric iris template. The proposed algorithm uses two separate keys: one derived from a logistic map equation and other from the linear feedback shift register. The two keys encrypt left iris and right iris images separately, thus giving an left encrypted image and right encrypted image. The two encrypted images are fused to get a final template to be stored. In order to evaluate and analyse the performance, the proposed algorithm was put through a cascade of tests: subjective and objective. Result analysis show that depict that key space is significantly enhanced in the proposed work. Also the proposed crypto-system shows good sensitivity to any change in the key.

The goal of this paper is to predict salary of a person after a certain year. The graphical representation of predicting salary is a process that aims for developing computerized system to maintain all the daily work of salary growth graph in any field and can predict salary after a certain time period. This application can take the database for the salary system from the organisation and makes a graph through this information from the database. It will check the salary fields then import a graph which helps to observe the graphical representation. And then it can predict a certain time period salary through the prediction algorithm. It can also be applied in some other effective prediction also.

This paper proposes a new way to measure the leverage effect and its propagation over time. We also show that, with respect to the newly proposed measure, common volatility models like the GJRGARCH, the Exponential GARCH, and the asymmetric SV can be inaccurate to correctly represent the leverage effect and its propagation for financial time series. We propose to modify the variance recursion of common volatility models by including an auxiliary leverage process which allows for a proper representation of the leverage effect and its propagation over time. Empirical results indicate that the inclusion of the auxiliary leverage process improves both in sample and out of sample.

This paper evaluates how adaptive learning agents weight different pieces of information when forming expectations with a recursive least squares algorithm. The analysis is based on a renewed and more general non-recursive representation of the learning algorithm, namely, a penalized weighted least squares estimator, where a penalty term accounts for the effects of the learning initials. The paper then draws behavioral implications of alternative specifications of the learning mechanism, such as the cases with decreasing, constant, regime-switching, adaptive, and age-dependent gains, as well as practical recommendations on their computation. One key new finding is that without a proper account for the uncertainty about the learning initial, a constant-gain can generate a time-varying profile of weights given to past observations, particularly distorting the estimation and behavioral interpretation of this mechanism in small samples of data. In fact, simulations and empirical estimation of a Phillips curve model with learning indicate that this particular misspecification of the initials can lead to estimates where inflation rates are less responsive to expectations and output gaps than in reality, or “flatter” Phillips curves.

This paper considers the characterization of the reduced-form solution of a large class of linear rational expectations models. I show that under certain conditions, if a solution exists and is unique, it can be cast in finite-order VAR form. I also investigate the conditions for the VAR form to be stationary with a well-defined residual variance-covariance matrix in equilibrium, for the shocks to be recoverable, and for local identification of the structural parameters for estimation from the sample likelihood. An application to the workhorse New Keynesian model with accompanying Matlab codes illustrates the practical use of the finite-order VAR representation. In particular, I argue that the identification of monetary policy shocks based on structural VARs can be more closely aligned with theory using the finite-order VAR form of the model solution characterized in this paper.

The cloud-computing technique provides services to the customer globally. The large-scale datacenters arrangement is very complex due to demand for optimization of resources utilization and power consumption. Simulation tools provide an easy way to model complex research problems through GUI representation. Cloudreports is a GUI based simulation tool that generates reports and provides an API with extensions and user defined plugins. In the present paper using this simulation tool user may set the simulation scenario for various VM’s and resources consumption, and simulation environment for the provider, customer, SAN, network performance etc. The resource utilization and overall resource utilization reports by various data centers, customers are also generated. PUE and DCIE, is also computed for effective power usage within datacenters.

We propose a novel estimation technique for a general approach to mixed frequency data which we label REtrieval from MIxed Sampling Frequency data (REMIS): Given data observed at mixed sampling frequency we aim to retrieve the parameters of the underlying high frequency possibly singular VAR system or Generalized Dynamic Factor model (GDFM) where the latent component has rational spectral density. A canonical state space representation is derived for the blocked process which second moments correspond to all those second moments observable under mixed frequency. This representation has nice features, it is mini phase and stable, minimal and many coordinates of the state are directly observed. Based on this representation a consistent subspace estimator for the high frequency parameters is derived. The advantage is that the directly observed components do not have to be estimated as opposed to other subspace procedures.

This paper investigates the role of technology levels in shaping the structure of the global value chain (GVC) at the macro level. We incorporate Ricardian comparative advantage into the production of intermediate goods involving both snake- and spider-type supply chains to capture the overall GVC integration. We analytically find that the country with a higher technology level produces the intermediate inputs at production stages involving a higher degree of difficulty, which is consistent with the real data. Furthermore, we verify how well the model fits the observed data by executing the calibration procedure using data from the World Input-Output Database. Our findings indicate strong correlations between calibrated outcomes and the observed data, as high as 87.3% for the GVC participation value and 79.3% for the real wage, proving that our model is a plausible representation of the structure of the GVC.

The present work proposes a methodology for the representation of performance scenario in Packaged Retail and Insurance-based Investment Products (PRIIPS), by the means of a no-arbitrage probability table easy to understand for the retail investor. A statistical reconstruction via the method of moments allows to capture the main properties of the PRIIP market implied distribution by identifying the minimum number of descriptive moments needed. A reasonable quantile partition that is effective for representing to the retail investor the complex distributions of structured products characterized by non-linear pay-offs is then proposed.

This paper introduces a new approach to face capital allocation problems from the perspective of acceptance sets, by defining the family of sub-acceptance sets. We study the relations between the notions of sub-acceptability and acceptability of a risky position as well as their impact on the allocation of risk. We define the notion of risk contribution rule and show how in this context it is interpretable as a tool for assessing the contribution of a sub-portfolio to a given portfolio in terms of acceptability without necessarily involving a risk measure. Furthermore, we investigate under which conditions on a risk contribution rule a representation of an acceptance set holds in terms of the risk contribution rule itself, thus extending to this setting the interpretation, classical in risk measures theory, of minimal amount required to hedge a risky position.

For the large family of ARMA models with variable coefficients we provide an explicit and computationally tractable solution representation, which yields the fundamental properties of such processes, including the Wold-Cramer decomposition and the covariance structure. These results are founded on a banded Hessenbergian representation of the Green's function, built up solely of the auto-regressive coefficients of the model. A generic condition, also expressed in terms of the Green's function,guarantees the convergence of the above mentioned properties. This condition is in line with the asymptotic stability and efficiency of such processes, while, their invertibility is guaranteed by an analogous condition, but now the Green's function is built up of the moving average coefficients. As a consequence, we illustrate mathematically a structural asymmetry between constant and 'time dependent' coefficient models, that is in the latter ones the backward and forward asymptotic efficiency differ structurally from one another. An alternative approach to the Hessenbergian solution representation is described by a simple procedure for manipulating polynomials with variable coefficients. The practical significance of the theoretical results in this work is illustrated with an application to U.S. inflation data. The main finding is that inflation persistence increased after 1976, whereas from 1986 on wards the persistence declines and stabilizes to even lower levels than the pre-1976 period.

This paper tackles the mixed-frequency modeling problem from a new perspective. Instead of drawing upon the common distributed lag polynomial model, we use a transfer function representation to develop a new type of models, named TF-MIDAS. We derive the theoretical TF-MIDAS implied by the high-frequency VARMA family models for two common aggregation schemes, flow and stock. This exact correspondence leads to potential gains in terms of now-casting and forecasting performance against the current alternatives. The estimation of the model proposed is also addressed via its state space equivalent form. A Monte Carlo simulation exercise confirms that TF-MIDAS beats U-MIDAS models (its natural competitor) in terms of out-of-sample now-casting performance for several data generating high-frequency processes.

This exploration project is a confirmatory analysis on many studies about classification of eggs to set the standard of classifying the eggs in the market. The study used Arduino Uno R2 kit with two devices namely a load cell (HX711) kit and the Pixy Camera to have variation in getting the data. The main goal of the study is to introduce new way of processing egg classifier. Specifically, to develop a project that classifies the three categories of the eggs, identify the different sizes and weight of the eggs and display the real-time number of classified eggs in a tabular and graphical representation in both desktop and Web application. The prototype was successfully designed with small changes from the materials and other positioning of devices because of the adjustment with the characteristics of the devices that are sensitive in nature. The major contribution of the study is the new way of processing that displays the sorted information based on the logic activity on classifying the eggs. A sophisticated desktop application used for collecting data and a functional website were the highlight of this study.

Numerical information can be represented in three formats: two symbolic (visual (digits) and verbal (number words)) and one nonsymbolic (analog) format. Studies have shown that the precision of symbolic numerical representation is associated with math performance. The precision of symbolic representation is mostly discussed as the precision of representation in a visual format, whereas the precision of representation in verbal format and its relation with math performance is less studied. The current study examines the precision of symbolic numerical representation in visual and verbal formats and the relationship between such precision and math performance when controlling for prior math performance, nonsymbolic numerical representation, phonological processing, reading skills and working memory. We used data from 367 Russian first graders (mean age, 7.6 years; 53% girls). To assess the precision of symbolic numerical representation, magnitude comparison tasks with digits and number words were used. It was found that the precision of symbolic representation in verbal format did not have a direct effect on math performance, but has an indirect effect via visual format of symbolic representation, even when controlling for prior math performance and other cognitive abilities.

When retaining a sequence of auditory tones in working memory (WM), two forms of information – frequency (content) and ordinal position (structure) – have to be maintained in the brain. Here, we employed a time-resolved multivariate decoding analysis to examine how content and structure are represented in human WM. We demonstrate that content and structure are stored in a dissociated manner and show distinct characteristics. First, each tone is associated with two separate codes in parallel, characterizing its frequency and ordinal position, respectively. Second, during retention, a structural retrocue reactivates structure but not content, whereas a white noise triggers content but not structure. Third, structure representation remains unchanged whereas content undergoes a transformation throughout memory progress. Finally, content and structure reactivations during retention correlate with WM behavior. Overall, our results support a factorized content-structure representation in auditory WM, which might help efficient memory formation by generalizing stable structure to new auditory inputs.

Genetic algorithms, computer programs that simulate natural evolution, are increasingly applied across many disciplines. They have been used to solve various optimisation problems from neural network architecture search to strategic games, and to model phenomena of adaptation and learning. Expertise on the qualities and drawbacks of this technique is largely scattered across the literature or former, motivating an compilation of this knowledge at the light of the most recent developments of the field. In this review, we present genetic algorithms, their qualities, limitations and challenges, as well as some future development perspectives. Genetic algorithms are capable of exploring large and complex spaces of possible solutions, to quickly locate promising elements, and provide an adequate modelling tool to describe evolutionary systems, from games to economies. They however suffer from high computation costs, difficult parameter configuration, and crucial representation of the solutions. Recent developments such as GPU, parallel and quantum computing, conception of powerful parameter control methods, and novel approaches in representation strategies, may be keys to overcome those limitations. This compiling review aims at informing practitioners and newcomers in the field alike in their genetic algorithm research, and at outlining promising avenues for future research. It highlights the potential for interdisciplinary research associating genetic algorithms to pulse original discoveries in social sciences, open ended evolution, artificial life and AI.

We address how recursive utility affects important results in the theory of economics of uncertainty and time, as compared to the standard model, where the focus is on dynamic models in discrete time. Several puzzles associated with the standard theory are less puzzling with recursive utility, even if this type of preference representation seems close to the standard one at first sight. An inconsistency with the axioms behind the standard, separable and additive expected utility representation is pointed out and extended to also be relevant for recursive utility. The basic difference from the standard model is that recursive utility allows a form of separation of consumption substitution from risk aversion. This also means that the timing of resolution of uncertainty matters. In dynamic models, however, this turns out to be a rather crucial step.

Current business models have struggled to support early-stage drug development. In this paper, we study an alternative financing model, the mega-fund structure, to fund drug discovery. We extend the framework proposed in previous studies to account for correlation between phase transitions in drug development projects, thus making the model a more realistic representation of bio-pharma research and development. In addition, we update the parameters used in our simulation with more recent estimates of the probability of success (PoS). We find that the performance of the mega-fund becomes less attractive when correlation between projects is introduced. However, the risk of default and the expected returns of the vanilla mega-fund remain promising even under moderate levels of correlation. In addition, we find that a leveraged mega-fund outperforms an equity-only structure over a wide range of assumptions about correlation and PoS.

We introduce a novel method to identify information networks in stock markets, which explicitly accounts for the impact of public information on investor trading decisions. We show that public information has a clear effect on the empirical investor networks’ topology. Most importantly, our method strengthens the identified relationship between investors’ network centrality and returns. Furthermore, when less significant links are re- moved, the association between centrality and returns becomes statistically and economically stronger. Findings suggest that our approach leads to a more precise representation of the information network. publishedVersion

We examine the complexity of financial returns generated by popular agent-based models through studying multifractal properties of such time series. Specifically, we are interested in the sensitivity of the models to their parameter settings and whether some patterns emerge in the connection between complexity and a specific type of parameter. We find that (i) herding behavior mostly boosts the model complexity as measured by multifractality, (ii) various in-built stabilizing factors increase model complexity, (iii) the role of the intensity of choice as well as the chartists' representation have rather model-specific effects, and (iv) the increasing number of agents has no statistically significant effect on the model complexity. Finally, the core feature driving the model complexity seems to be the implementation of a switching mechanism governing agents' interactions. The heterogeneous set of nine analyzed models thus offers some universal concepts that hold across their range. Our results also indicate that complex dynamics are observed not only for the benchmark parameter settings but also for other combinations of parameter values for most models. This opens new avenues for future research and specifically motivates examining the models in more detail by focusing on other dynamic properties in addition to the herein presented multifractality.

We investigate the hot hand effect in the game of cricket by analyzing the complete recorded history of international cricket. We introduce an original temporal representation of performance streaks, which is suitable to be modelled as a self-exciting point process. We confirm the presence of hot hands across the players' careers. We show that the self-excitation patterns in performance clusters can be exploited for predicting future performances. This paper contributes to recent historiographical debates concerning the presence of hot hands in the sequence of successes in individual performances. The introduction of several metrics and methods can be useful to test and exploit the clustering of performance in the study of human behavior and the design of algorithms for predicting success.

Summary To formulate a machine learning (ML) model to establish the polymer's structure-property correlation for glass transition temperature Tg, we collect a diverse set of nearly 13,000 real homopolymers from the largest polymer database, PoLyInfo. We train the deep neural network (DNN) model with 6,923 experimental Tg values using Morgan fingerprint representations of chemical structures for these polymers. Interestingly, the trained DNN model can reasonably predict the unknown Tg values of polymers with distinct molecular structures, in comparison with molecular dynamics simulations and experimental results. With the validated transferability and generalization ability, the ML model is utilized for high-throughput screening of nearly one million hypothetical polymers. We identify more than 65,000 promising candidates with Tg > 200°C, which is 30 times more than existing known high-temperature polymers (∼2,000 from PoLyInfo). The discovery of this large number of promising candidates will be of significant interest in the development and design of high-temperature polymers., Graphical abstract, Highlights • Large datasets for polymer's glass transition temperature are collected • Transferability of ML models depends on feature representations • Molecular dynamics models and experimental results validate the formulated ML model • Extensive promising candidates for high-temperature polymers are screened by ML model, The bigger picture The design and development of high-temperature polymers has been an experimentally driven and trial-and-error process guided by experience, intuition, and conceptual insights. However, such an Edisonian approach is often costly, slow, biased toward certain chemical space domains, and limited to relatively small-scale studies, which may easily miss promising compounds. To overcome this challenge, we formulate a data-driven machine learning (ML) approach, integrated with high-fidelity molecular dynamics simulations, for quantitatively predicting the glass transition temperature of a polymer from its chemical structure and rapid screening of promising candidates for high-temperature polymers. Our work demonstrates that ML is a powerful method for the prediction and rapid screening of high-temperature polymers, particularly with growing large sets of experimental and computational data for polymeric materials., Polymers with outstanding high-temperature properties have been identified as promising materials for aerospace, electronics, and automotive applications. However, the current design and development of high-temperature polymers has been an experimentally driven and trial-and-error process guided by experience, intuition, and conceptual insights. Therefore, we formulate a machine learning model that can quantitatively predict the glass transition temperature of a polymer from its chemical structure, such that more promising high-temperature polymers can be efficiently filtered out through high-throughput screening.

In this paper I evaluate the Journal Impact Factor using a theory of measurement. I argue that JIF does not stand up to close scrutiny. To measure a concept adequately, our theory of measurement requires correspondence between three steps: the characterisation of the concept, its representation, and the procedures followed to carry out the measurement. Characterisation involves defining the concept: identifying its boundaries, which fixes the features that belong to it. Representation involves defining a metrical system that appropriately represents the concept. The procedures are the rules formulated for applying the metrical system to the tokens. These three steps do not line up together neatly for JIF. There are at least two problems. First, the procedures to measure JIF do not reflect an unequivocal characterisation. Second, the representation strategy of JIF is inappropriate and not justified, given the kind of concept it tries to capture: one without strict boundaries. The bottom line is not that the JIF ought to be eschewed. Sufficient reasons related to how JIF distorts scientists’ incentives have been provided to this end. But path dependence is a tricky issue—the longevity of qwerty keyboards demonstrates it. The bottom line is that, given that JIF is unlikely to vanish, we better start giving it some proper scientific basis.

The principle of indifference states that events should be assigned equal probabilities if no reason can be given for regarding one event as more likely than another. This paper provides a normative argument for the principle of indifference based on a new formal model of decision under uncertainty, called naive uncertainty. The model consists of a set of preferences and their utility representation, from which the principle of indifference arises as utility maximizing. Faced with a finite collection of choice alternatives to which allocation weights are to be assigned, the decision maker under naive uncertainty essentially considers all possible probability distributions associated with the random payoffs, and chooses the allocation that minimizes the variability of outcome across all probabilities. The optimal allocation under naive uncertainty is shown to be the equal weighted allocation. Diverse examples of applications of the model of naive uncertainty and the principle of indifference are given, including naive diversification, evolutionary behavior, and policy making using expert panels.

Importance of distributional channels in macroeconomic dynamics has been the object of considerable attention in empirical studies. Despite significant amount of effort aimed at incorporating heterogeneity into macroeconomics, however, their explicit inclusion in the standard policy toolbox is far from widespread. A relevant obstacle, in such cases, is the computation of equilibria. I propose a global solution method for the computation of infinite-horizon, heterogeneous agent macroeconomic models with aggregate uncertainty. Details of the algorithm are illustrated by presenting its application to a an example model: in it, aggregate dynamics depends explicitly on firm entry and exit, and individual choices are often constrained by a form of market incompleteness. Existing computational strategies are either unfeasible or provide inaccurate solutions. Moreover, global solutions are computationally expensive because the minimal representation of the aggregate state space - and thus the aggregate law of motion - faces the curse of dimensionality. The proposed strategy thus combines adaptive sparse grids with a cross-sectional density approximation, and introduces a framework for solving the more general class of dynamic models with firm or household heterogeneity accurately.

The continuum hypothesis has been unsolved for hundreds of years. That said, can it be completely answered? With disproving the cultural competent continuum (Lam, 2014), one can link it with the mathematical continuum and possibly disprove the continuum hypothesis problem (Lam, 2018). It may also be possible to find out the conditions of its independence to ZF(C). To move forward, one may extend our mathematical system (by employing a stronger set theory) and solve the continuum problem. This is similar to three cases in the discriminant of solving quadratic equations. Hence, an algorithmic flowchart can completely solve and depict the problem. From the above case, one may state the following: when mathematics is extended into a new system, important mathematical issues can be solved. Indeed, there are different types of such mathematical systems — similar to ancient mathematical numeral symbols. Hence, different cultures have different ways of representation, which is similar to a Chinese saying: “different villages have different laws”. However, the primary purpose of mathematical symbols was initially for memorisation and communication. This implies that the basic purpose for developing any new mathematical system is to help solve natural phenomenon in our universe.

Representative models of the macroeconomy (RMs), such as DSGE models, frequently contain unobserved variables. A finite-order VAR representation in the observed variables may not exist, and therefore the impulse responses of the RMs and SVAR models may differ. We demonstrate this divergence often is: (i) not substantial; (ii) reflects the omission of stock variables from the VAR; and (iii) when the RM features I (1) variables can be ameliorated by estimating a latent-variable VECM. We show that DSGE models utilize identifying restrictions stemming from common factor dynamics reflecting statistical, not economic, assumptions. We analyze the use of measurement error, and demonstrate that it may result in unintended consequences, particularly in models featuring I (1) variables.

We propose a theory for rating financial securities based on a concept of self-consistency, which does not allow issuers to gain, by tranching financial securities, from investors who rely on the rating criterion. While the expected loss criterion used by Moody's satisfies self-consistency, the probability of default criterion used by S&P does not. We find empirical evidences in the post-Dodd-Frank period (i.e., after July 2010) that the issuers may take advantage of the absence of self-consistency. We further propose a concept of scenario-relevance which reflects practical evaluation procedures of potential losses from defaultable securities. Our main theoretical results show that a self-consistent rating measure admits a Choquet integral representation, and this representation is also analytically tractable if one further takes economic scenarios into account. We suggest new examples of self-consistent and scenario-based rating criteria, such as ones based on the VaR and the Expected Shortfall.

Algebraic calculation of the fiscal multiplier ignores the concept of velocity of money. Here, we incorporate the concept of velocity of money in the algebraic derivation of the fiscal multiplier which results into a slightly different representation. Then, we empirically calculate the values of fiscal multipliers for 05 (five) OECD countries using our proposed algebraic representation and compare the results with the classical Keynesian values. Our results also point out why the theoretical values of the fiscal multipliers are higher than the empirical values.

We propose a multiple-prior model of preferences under ambiguity that provides a unified lens through which to understand different formalizations of ambiguity aversion, as well as context-dependent negative and positive ambiguity attitudes documented in experiments. This model, Boolean expected utility (BEU), represents the belief the decision-maker uses to evaluate any uncertain prospect as the outcome of a game between two conflicting forces, Pessimism and Optimism. We prove, first, that BEU provides a novel representation of the class of invariant biseparable preferences (Ghirardato, Maccheroni, and Marinacci, 2004). Second, BEU accommodates rich patterns of ambiguity attitudes, which we characterize in terms of the relative power allocated to each force in the game.