Loss Function Approach Research Articles

An efficient loss function and deep learning approach for ranking stock returns in the absence of prior knowledge

To pursue profit from the dynamic, complex, and noisy stock markets, various efforts utilizing deep learning methods to forecast asset price movements have sprung up. We observe that there are two issues in the current work. Firstly, there exists a discrepancy between the forecasting target and actual profitability, as achieving better forecasting results does not necessarily guarantee higher profits. Secondly, many existing methods heavily rely on prior knowledge during the forecasting process, which entails the need for information gathering and may not adapt well to dynamic and complex market conditions. For the first issue, we design a novel reward learning loss function as an optimization object to better model the bridge between the forecasting target and the profit from the trading process. To solve the second issue, we propose a structure based on multi-head attention to model the inter-stock relations directly from trading data without relying on prior knowledge. Also, we present a simple time-asynchronous attention-based method to model the lead–lag phenomenon in the market. We conduct experiments using over 600 stocks of the CSI100, CSI300, and CSI500 indexes from 2010 to 2020 with five strong baselines. The experimental results demonstrate that our methods achieve annualized returns of 5%, 10%, and 13% for long and 5%, 6%, and 8% for short above the optimal baseline results on the three indexes. Further analysis shows that our RL-Loss is better than classic PR-Loss, and the inter-stock relation modeling methods proposed without prior knowledge are effective.

Prediction techniques for the scaled models made of different materials under impact loading

Predicting the impact response of the full-size structure by scaled models is of great importance in practice. The traditional similarity relationships will not be satisfied when the material of scaled models differs from that of the prototype. A dynamic similarity method is proposed to predict the dynamic behavior of the full-size structure under impact loading, where the strain hardening and the strain rate effects of different materials are both considered. In this work, we firstly develop the loss function to quantify the flow yield stress difference between the scaled model and the prototype during the impact, further demonstrating that the optimal velocity scaling factor can be obtained when the total difference is minimum. A simple and practical material selection criterion is also proposed to guide the selection of the most suitable alternative material for the scaled model. Two impacted structures are verified numerically, the results show that scaled models made of different materials produce small errors in both time and space fields of collision force and displacement when the proposed loss function approach is used to predict the prototype response. It is also confirmed that the proposed material selection criteria can foresee the scaled model made of which material can achieve higher prediction accuracy.

Model-Robust Estimation of Multiple-Group Structural Equation Models

Structural equation models (SEM) are widely used in the social sciences. They model the relationships between latent variables in structural models, while defining the latent variables by observed variables in measurement models. Frequently, it is of interest to compare particular parameters in an SEM as a function of a discrete grouping variable. Multiple-group SEM is employed to compare structural relationships between groups. In this article, estimation approaches for the multiple-group are reviewed. We focus on comparing different estimation strategies in the presence of local model misspecifications (i.e., model errors). In detail, maximum likelihood and weighted least-squares estimation approaches are compared with a newly proposed robust Lp loss function and regularized maximum likelihood estimation. The latter methods are referred to as model-robust estimators because they show some resistance to model errors. In particular, we focus on the performance of the different estimators in the presence of unmodelled residual error correlations and measurement noninvariance (i.e., group-specific item intercepts). The performance of the different estimators is compared in two simulation studies and an empirical example. It turned out that the robust loss function approach is computationally much less demanding than regularized maximum likelihood estimation but resulted in similar statistical performance.

The Interaction of OTUB1 and TRAF3 Mediates NLRP3 Inflammasome Activity to Regulate TGF-β1-induced BEAS-2B Cell Injury.

Asthma is a frequently chronic respiratory disease with inflammation and remodeling in the airway. OTUB1 has been reported to be associated with pulmonary diseases. However, the role and potential mechanism of OTUB1 in asthma remain unclear. The expressions of OTUB1 in the bronchial mucosal tissues of asthmatic children and TGF-β1-induced BEAS-2B cells were determined. The biological behaviors were assessed in an asthma in vitro model using a loss-function approach. The contents of inflammatory cytokines were detected by ELISA kits. The related protein expressions were performed using western blot assay. Besides, the interaction between OTUB1 and TRAF3 was detected by Co-IP and ubiquitination assays. Our results showed that OTUB1 level was increased in asthmatic bronchial mucosal tissues and TGF-β1-induced BEAS-2B cells. OTUB1 knockdown promoted proliferation, inhibited apoptosis and EMT of TGF-β1-treated cells. The inhibition of OTUB1 attenuated the TGF-β1-induced inflammation and remodeling. Furthermore, OTUB1 knockdown inhibited the deubiquitination of TRAF3 and further suppressed the activation of NLRP3 inflammasome. The overexpression of TRAF3 or NLRP3 reversed the positive role of OTUB1 knockdown in TGF-β1-induced cells injury. Collectively, OTUB1 deubiquitinates TRAF3 to activate NLRP3 inflammasome, thereby leading to inflammation and remodeling of TGF-β1-induced cells, and further promoting the pathogenesis of asthma.

Deep-learning-based fault detection and recipe optimization for a plastic injection molding process under the class-imbalance problem

AbstractThis paper proposes a supervised learning with a class-balancing loss function (SL-CBL) approach for fault detection and feature-similarity-based recipe optimization (FSRO) for a plastic injection molding process. SL-CBL is a novel method that can accurately classify an input sample as a normal or fault condition, even when the training data are severely class-imbalanced. The proposed class-balancing loss function consists of the weighted focal loss and the loss of the F1 score; together, these are used to correctly classify even a small number of faulty samples. SL-CBL is investigated with four classifiers of different structures; the classifiers consist of several fully connected and batch normalization layers. FSRO is an optimization scheme that finds the optimal recipe whose feature is similar to the features of normal samples. The optimal solution is obtained by minimizing the Euclidean distance to the centroid of the normal features. In this research, the proposed SL-CBL and FSRO methods are validated by applying them to an industrial plastic injection molding dataset. The validation results show that the proposed SL-CBL approach achieves the highest F1 score with the lowest misclassification rate, as compared to the alternative methods. When visualizing the feature space, the optimal recipe found by the FSRO scheme was found to be close to the centroid of the normal features, even if the initial recipe is classified as a fault. Furthermore, each variable of the optimized recipe lies within the confidence interval of 3${\rm{\sigma }}$ for the normal condition. This indicates that the optimal recipe is statistically similar to the normal samples.

Segmentation with Residual Attention U-Net and an Edge-Enhancement Approach Preserves Cell Shape Features.

The ability to extrapolate gene expression dynamics in living single cells requires robust cell segmentation, and one of the challenges is the amorphous or irregularly shaped cell boundaries. To address this issue, we modified the U-Net architecture to segment cells in fluorescence widefield microscopy images and quantitatively evaluated its performance. We also proposed a novel loss function approach that emphasizes the segmentation accuracy on cell boundaries and encourages shape feature preservation. With a 97% sensitivity, 93% specificity, 91% Jaccard similarity, and 95% Dice coefficient, our proposed method called Residual Attention U-Net with edge-enhancement surpassed the state-of-the-art U-Net in segmentation performance as evaluated by the traditional metrics. More remarkably, the same proposed candidate also performed the best in terms of the preservation of valuable shape features, namely area, eccentricity, major axis length, solidity and orientation. These improvements on shape feature preservation can serve as useful assets for downstream cell tracking and quantification of changes in cell statistics or features over time.

Optimizing of the Parameters of Coconut Sugar Production Using Taguchi Design in Riau, Indonesia

The quality of coconut sugar produced from Indragiri Hilir Regency, Riau Province, is still fluctuating and has not compliant with SNI-01-3743 of 1995, which incorrect parameters in its production may cause incorrect parameters in its production. Optimizing of production parameters is needed to maintain the quality of coconut sugar. Coconut sugar quality indicators can be determined from the physical-chemical and organoleptic characteristics. Therefore, it is necessary to modify the parameters of coconut sugar production, aiming to obtain optimal production parameters to produce a standard production process and coconut sugar following consumer needs and SNI. The experimental design uses the Taguchi method, which consists of 4- factor variables: (1) the type of sap preservative, (2) defoaming agent, (3) the time of tapping, and (4) the cooking temperature of the sap. The multi-response loss function approach was used to select the best production process parameters based on the typical characteristics of qualities. The results showed that the optimal production parameters were obtained in a combination of mangosteen peel and lime as sap preservative, coconut milk as a defoaming agent, long-time tapping time of 8 hours (heating) +16 hours, and cooking temperature of 135oC with a yield of 15.58%, part of water-insoluble in 0.07%, (ΔE) color 62.30, hardness 99.92 N, moisture content 6.87%, ash content 1.96%, sucrose content 77.77%, reducing sugar content 8.09% with organoleptic test values close to 4 ( like) and equal SNI.

Integration of Multivariate Loss Function Approach in the Hotelling’s Charts under Banerjee-Rahim (1988) Weibull Shock Models

A proper monitoring of stochastic systems is the control charts of statistical process control and drift in characteristics of output may be due to one or several assignable causes. Although many research works have been done on the economic design of control charts with single assignable cause, the economic statistical design of T^2 control chart under Weibull shock model with multiple assignable causes and considering multivariate Taguchi loss function has not been presented yet. Using Taguchi loss function in the concept of quality control charts with economic and economic statistical design leads to better decisions in the industry. Based on the optimization of the average cost per unit of time and taking into account the different combination values of Weibull distribution parameters, optimal design values ??of sample size, sampling interval and control limit coefficient were derived and calculated. Then the cost models under non-uniform and uniform sampling scheme were compared. The results revealed that the model under multiple assignable causes with Taguchi loss function has a lower cost than single assignable cause model and integrated model with non-uniform sampling has a lower cost than that with uniform sampling.

Multivariate-Based Technique for Solving Multi-Response Surface Optimization (MRSO) Problems: The Case of a Maximization Problem

Multi-response surface optimization (MRSO) is a problem that is peculiar to an industrial setting, where the aim of a process engineer is to set his process at operating conditions that simultaneously optimize a set of process responses. In Statistics, several methods have been proffered for tackling problems of this nature. Some of such methods are that of: overlapping contour plots, constrained optimization problem, loss function approach, process capability approach, distance function approach, game theory approach, and the desirability function approach. These, methods are however, not without teething flaws as they are either too problem specific, or require very complex and inflexible routines; little wonder, the method of desirability function has gained popularity especially because it overcomes the latter limitation. In this article, we have proposed and implemented a multivariate-based technique for solving MRSO problems. The technique fused the ideas of response surface methodology (RSM), multivariate multiple regression and Pareto optimality. In our technique, RSM was implemented on an all-maximization problem as a case-study process; in which case, first-order models (FOMs) for the responses were fitted using 2k factorial designs until the FOMs proved to be inadequate, while uniform precision rotatable central composite design was used to obtain second-order models (SOMs) for the respective responses in the event of model inadequacy of the FOMs. With the implementation of the proposed technique to the case study, optimal operating conditions were obtained, with observations stemming thereof summarized as axioms. The first, second and third axioms respectively stated that: (1) the mid-point of all optimal operating conditions obtained via the proposed technique is Pareto optimal, (2) the mid-point of all optimal responses at the Pareto optimal operating condition is Pareto optimal, and (3) the region bounded by each of the optimal operating conditions from each second-order model (SOM) is a Pareto front.

Pricing real options based on linear loss functions and conditional value at risk

The main purpose of this paper is to expand real option analysis out of the realm of pure financial option pricing techniques. To overcome many of the well-known concerns by adopting the financial option pricing techniques for modeling real options problems such as replicating portfolio concept, geometric Brownian motion as underlying stochastic process, and estimating project volatility, we propose an alternative real option valuation based on the loss function approach. The option value determined by the loss function approach is equivalent to the expected value of perfect information (EVPI) in decision analysis. It basically sets the upper bound of risk premium to pay in retaining the options. In practice, many firms utilize the concept of Value at Risk to manage their portfolio risk. If a firm sets a target VAR, then we may be able to link this VAR in refining the actual risk premium to pay in hedging the risk embedded in the investment. With this practice in mind, we present a logic to figure out an appropriate amount of real option premium to pay for a given level of risk tolerance. A comprehensive example is presented to demonstrate the computational procedures as well as economic interpretations on the outcomes.

Automated diagnosis of COVID-19 with limited posteroanterior chest X-ray images using fine-tuned deep neural networks.

The novel coronavirus 2019 (COVID-19) is a respiratory syndrome that resembles pneumonia. The current diagnostic procedure of COVID-19 follows reverse-transcriptase polymerase chain reaction (RT-PCR) based approach which however is less sensitive to identify the virus at the initial stage. Hence, a more robust and alternate diagnosis technique is desirable. Recently, with the release of publicly available datasets of corona positive patients comprising of computed tomography (CT) and chest X-ray (CXR) imaging; scientists, researchers and healthcare experts are contributing for faster and automated diagnosis of COVID-19 by identifying pulmonary infections using deep learning approaches to achieve better cure and treatment. These datasets have limited samples concerned with the positive COVID-19 cases, which raise the challenge for unbiased learning. Following from this context, this article presents the random oversampling and weighted class loss function approach for unbiased fine-tuned learning (transfer learning) in various state-of-the-art deep learning approaches such as baseline ResNet, Inception-v3, Inception ResNet-v2, DenseNet169, and NASNetLarge to perform binary classification (as normal and COVID-19 cases) and also multi-class classification (as COVID-19, pneumonia, and normal case) of posteroanterior CXR images. Accuracy, precision, recall, loss, and area under the curve (AUC) are utilized to evaluate the performance of the models. Considering the experimental results, the performance of each model is scenario dependent; however, NASNetLarge displayed better scores in contrast to other architectures, which is further compared with other recently proposed approaches. This article also added the visual explanation to illustrate the basis of model classification and perception of COVID-19 in CXR images.

Integration of quadratic loss function approach in the economic design of VSI control chart with combined attribute‐variable inspection

AbstractThis paper proposes a variable sampling interval control chart for monitoring the process mean using combined attribute‐variable inspection. An economic design of the proposed chart using Taguchi's quadratic loss function is presented. The proposed model minimizes the total expected cost per unit time while minimizing out‐of‐control average time‐to‐signal and maintaining reasonable in‐control average time‐to‐signal. The numerical results indicate that, compared with the corresponding fixed sampling interval control chart scheme, the proposed control chart yields better performance in terms of both shorter time to signal a process shift and less total expected cost per unit time. Sensitivity analyses of the cost model are also carried out by using a fractional factorial experimental design.

Design Approach for Quality Improvement and Cost Reduction: A Case Study of Cable Manufacturing

In this study, an extrusion process was studied in a cable manufacturing company located in Southeastern Nigeria. The critical-to-quality (CTQ) characteristic considered in the study was the cable diameter and the design effort was to reduce the possibilities of having cables with inconsistent dimensions. Taguchi Orthogonal Array Design L16 (4˄2) was used to achieve the optimum parameter settings after statistically investigating the assumed correlations between the process parameters in relation to cable dimension. Thereafter, an appropriate engineering tolerance interval for the improved process was designed to tighten the existing tolerance of the process from T±0.185 to T±0.032, thus reducing the engineering tolerance by 82.7% from the initial tolerance limit. After the process improvement, the Taguchi loss function approach was used to estimate cost attributed to deviations. The loss function estimation result has shown that the cost attributed to cable diameter deviations from the nominal value reduced from the initial value of ₦7.34/coil to a reduced cost of ₦2.08/coil. The sigma level increased from 0.6 to 5.2 at the project termination stage and the estimated annual loss decreased by 72% from the baseline value.

An early-stage classification of Lung Nodules by an Android based application using Deep Convolution Neural Network with Cost-sensitive loss function and Progressive scaling approach

An early-stage classification of Lung Nodules by an Android based application using Deep Convolution Neural Network with Cost-sensitive loss function and Progressive scaling approach

Robust parameter design based on Gaussian process with model uncertainty

In robust parameter design, it is common to use computer models to simulate the relationships between input variables and output responses. However, for the contaminated experimental data, the model uncertainty between computer models and actual physical systems will seriously impair the robustness of the optimal input settings. In this paper, we propose a new weighted robust design approach concerning the model uncertainty from outliers based on the robust Gaussian process model with a Student-t likelihood (StGP). Firstly, to reduce the impact of outliers on the output means and variances, the StGP modelling technique is adopted to estimate the relationship models for contaminated data. Secondly, the Gibbs sampling technique is employed to estimate model parameters for better mixing and convergence. Finally, an optimisation scheme integrating the quality loss function and confidence interval analysis approach is built to find the feasible optimisation solution. Meanwhile, the hypersphere decomposition method and data-driven method are applied to determine the relative weights of objective functions. Two examples are used to demonstrate the effectiveness of the proposed approach. The comparison results show that the proposed approach can achieve better performance than other approaches by considering the model uncertainty from outliers.

MicroRNA-495 confers inhibitory effects on cancer stem cells in oral squamous cell carcinoma through the HOXC6-mediated TGF-\u03b2 signaling pathway

BackgroundOral squamous cell carcinoma (OSCC) is associated with high morbidity and ranks sixth among malignancies worldwide. Increasing evidence suggests that microRNAs (miRNAs or miRs) play a critical role in regulating cancer stem cells (CSCs), which drive the proliferation and spread of OSCC. Therefore, based on the alteration of aberrantly expressed miR-495 and homeobox C6 (HOXC6) by Gene Expression Omnibus (GEO) analysis, we subsequently explore the potential effect of miR-495 on the progression of CSCs in OSCC.MethodsAfter the isolation of CSCs from the clinical tissue samples of OSCC patients, the expression of miR-495 and HOXC6 was determined, followed by the validation of the relationship between miR-495 and HOXC6. Subsequently, gain- and loss-function approach was performed to detect the role of miR-495 and HOXC6 in cell proliferation, migration, invasion, cell cycle entry, apoptosis, and epithelial-mesenchymal transition (EMT) of CSCs in OSCC, as well as the tumor growth in vivo.ResultsHOXC6 was highly expressed while miR-495 was poorly expressed in OSCC. HOXC6 was verified to be a target gene of miR-495, and miR-495 could inhibit the activation of the TGF-β signaling pathway. CSCs with miR-495 overexpression or HOXC6 silencing exhibited reversed EMT process; reduced abilities of proliferation, migration, and invasion; and promoted cell apoptosis in vitro. Moreover, inhibited tumor growth was observed in vivo after injection with miR-495 agomir or sh-HOXC6. In contrast, the downregulation of miR-495 showed an induced role in the progression of OSCC.ConclusionThese findings suggest that miR-495 may suppress HOXC6 to inhibit EMT, proliferation, migration, and invasion while promoting apoptosis of CSCs in OSCC by inhibiting the TGF-β signaling pathway.

Revealing forecaster's preferences: A Bayesian multivariate loss function approach

AbstractRevealing the underlying preferences of a forecaster has always been at the core of much controversy. Herein, we build on the multivariate loss function concept and propose a flexible and multivariate family of likelihoods. This allows examining whether a vector of forecast errors, along with control variables, shapes a forecaster's preferences and, therefore, the underlying multivariate, nonseparable, loss function. We estimate the likelihood function using Bayesian exponentially tilted empirical likelihood, which reveals the shape of the parameter and the power of the multivariate loss function. In the empirical section, the reported evidence reveals that the EU Commission forecasts are predominantly asymmetric, leaning towards optimism in the year ahead, while a correction towards pessimism occurs in the current year forecast. There is some variability of this asymmetry across member states, with forecasts, i.e. gross domestic product growth, for large Member States exhibiting more optimism

Actors, actions, and uncertainties: optimizing decision-making based on 3-D structural geological models

Abstract. Uncertainties are common in geological models and have a considerable impact on model interpretations and subsequent decision-making. This is of particular significance for high-risk, high-reward sectors. Recent advances allows us to view geological modeling as a statistical problem that we can address with probabilistic methods. Using stochastic simulations and Bayesian inference, uncertainties can be quantified and reduced by incorporating additional geological information. In this work, we propose custom loss functions as a decision-making tool that builds upon such probabilistic approaches. As an example, we devise a case in which the decision problem is one of estimating the uncertain economic value of a potential fluid reservoir. For subsequent true value estimation, we design a case-specific loss function to reflect not only the decision-making environment, but also the preferences of differently risk-inclined decision makers. Based on this function, optimizing for expected loss returns an actor's best estimate to base decision-making on, given a probability distribution for the uncertain parameter of interest. We apply the customized loss function in the context of a case study featuring a synthetic 3-D structural geological model. A set of probability distributions for the maximum trap volume as the parameter of interest is generated via stochastic simulations. These represent different information scenarios to test the loss function approach for decision-making. Our results show that the optimizing estimators shift according to the characteristics of the underlying distribution. While overall variation leads to separation, risk-averse and risk-friendly decisions converge in the decision space and decrease in expected loss given narrower distributions. We thus consider the degree of decision convergence to be a measure for the state of knowledge and its inherent uncertainty at the moment of decision-making. This decisive uncertainty does not change in alignment with model uncertainty but depends on alterations of critical parameters and respective interdependencies, in particular relating to seal reliability. Additionally, actors are affected differently by adding new information to the model, depending on their risk affinity. It is therefore important to identify the model parameters that are most influential for the final decision in order to optimize the decision-making process.

Conceptual Issues in Calibrating the Basel III Countercyclical Capital Buffer

This paper discusses issues in calibrating the countercyclical capital buffer (CCB) based on a sample of EU countries. It argues that the main indicator for buffer decisions under the Basel III framework, the credit-to-GDP gap, does not always work best in terms of covering bank loan losses that go beyond what could be expected from economic downturns. Instead, in the case of countries with short financial cycles and/or low financial deepening such as transition and developing economies, the Basel gap is shown to work best when computed with a low, smoothing factor and adjusted for the degree of financial deepening. The paper also analyzes issues in calibrating an appropriate size of the CCB and, using a loss function approach, points to a tradeoff between stability of the buffer size and cost efficiency considerations.

Conceptual Issues in Calibrating the Basel III Countercyclical Capital Buffer

This paper discusses issues in calibrating the countercyclical capital buffer (CCB) based on a sample of EU countries. It argues that the main indicator for buffer decisions under the Basel III framework, the credit-to-GDP gap, does not always work best in terms of covering bank loan losses that go beyond what could be expected from economic downturns. Instead, in the case of countries with short financial cycles and/or low financial deepening such as transition and developing economies, the Basel gap is shown to work best when computed with a low, smoothing factor and adjusted for the degree of financial deepening. The paper also analyzes issues in calibrating an appropriate size of the CCB and, using a loss function approach, points to a tradeoff between stability of the buffer size and cost efficiency considerations.