Two-step Selection Procedure Research Articles

Improving Cointegration-Based Pairs Trading Strategy with Asymptotic Analyses and Convergence Rate Filters

A pairs trading strategy (PTS) constructs a mean-reverting portfolio whose logarithmic value moves back and forth around a mean price level. It makes profits by longing (or shorting) the portfolio when it is underpriced (overpriced) and closing the portfolio when its value converges to the mean price level. The cointegration-based PTS literature uses the historical sample mean and variance to establish their open/close thresholds, which results in bias thresholds and less converged trades. We derive the asymptotic mean around which the portfolio value oscillates. Revised open/close thresholds determined by our asymptotic mean and standard derivations significantly improve PTS performance. The derivations of asymptotic means can be extended to construct a convergence rate filter mechanism to remove stock pairs that are unlikely to be profitable from trading to further reduce trading risks. Moreover, the PTS literature oversimplifies the joint problem of examining a stock pair’s cointegration property and selecting the fittest vector error correction model (VECM). We propose a two-step model selection procedure to determine the cointegration rank and the fittest VECM via the trace and likelihood ratio tests. We also derive an approximate simple integral trading volume ratio to meet no-odd-lot trading constraints. Experiments from Yuanta/P-shares Taiwan Top 50 Exchange Traded Fund and Yuanta/P-shares Taiwan Mid-Cap 100 Exchange Traded Fund constituent stock tick-by-tick backtesting during 2015–2018 show remarkable improvements by adopting our approaches.

Quantifying time perception during virtual reality gameplay using a multimodal biosensor-instrumented headset: a feasibility study.

We have all experienced the sense of time slowing down when we are bored or speeding up when we are focused, engaged, or excited about a task. In virtual reality (VR), perception of time can be a key aspect related to flow, immersion, engagement, and ultimately, to overall quality of experience. While several studies have explored changes in time perception using questionnaires, limited studies have attempted to characterize them objectively. In this paper, we propose the use of a multimodal biosensor-embedded VR headset capable of measuring electroencephalography (EEG), electrooculography (EOG), electrocardiography (ECG), and head movement data while the user is immersed in a virtual environment. Eight gamers were recruited to play a commercial action game comprised of puzzle-solving tasks and first-person shooting and combat. After gameplay, ratings were given across multiple dimensions, including (1) the perception of time flowing differently than usual and (2) the gamers losing sense of time. Several features were extracted from the biosignals, ranked based on a two-step feature selection procedure, and then mapped to a predicted time perception rating using a Gaussian process regressor. Top features were found to come from the four signal modalities and the two regressors, one for each time perception scale, were shown to achieve results significantly better than chance. An in-depth analysis of the top features is presented with the hope that the insights can be used to inform the design of more engaging and immersive VR experiences.

A two-step feature selection procedure for relevant markers of Squamous Cell Lung Carcinoma using different survival models

There are potentially infinite gene expression markers for Lung Squamous Cell Carcinoma. This results in a high-dimensional data with a large number of features. The selection of relevant markers for analysis is thus, of utmost importance. In our study, we have aimed to select a subset of prominent and significant features from 31918 features of gene expressions. Analysis is then performed on the selected features using the Cox Proportional Hazards Model to know how each marker affects the survival estimates of a patient. We have employed a two-step selection process to select a subset of markers. The first step is done by L1 regularized Cox PH. Then the selected markers are screened a second time by running a univariate Cox PH model and checking for the p-value of each bio-marker via Wald inference (p<0.05). Once the final selection is made, we estimate the Hazard Ratio and Confidence intervals using Maximum Likelihood Estimates (MLE) and the Bayesian Approach with the Cox Proportional Hazards Model (CPH) and the Accelerated Failure Time Model (AFT) as an alternative. A forest plot has also been generated to show the graphical representation of the meta-analysis done in the study. With the proposed selection procedure we have managed to find a suitable subset out of a large number of variables available. The features selected have been analyzed and their validity has been confirmed by using survival models.

Does urban-rural disparity exist in nicotine and alcohol dependence among Chinese older adults?

Background China is a large country with substantial urban-rural disparity. With the increasing concerns related to the rapid growth of the older adult population and problems related to smoking and alcohol consumption, this study aims to examine the potential urban-rural disparity associated with nicotine or alcohol dependence among Chinese older adults. Methods We used three waves of the Chinese Longitudinal Healthy Longevity Survey and included older adults who were 65 years old or above (CLHLS; n = 18,207). The Heckman two-step selection procedure was applied to reduce potential selection bias. The first and the second steps of the Heckman two-step selection procedure all included multivariable logistic regressions. Results In the final study sample, approximately 8.1% and 4.0% of older adults reported nicotine and alcohol dependence, respectively. In the first step of the Heckman selection procedure, urban residents were less likely to become current smokers and alcohol users than rural residents (all p < .05). However, urban-rural disparity was not associated with either nicotine or alcohol dependence (all p > .05). Conclusions We did not observe the urban-rural disparity in nicotine and alcohol dependence among Chinese older adults. Chinese policymakers should continue to strengthen national policy to combat smoking and alcohol consumption, especially older adults.

Employability and Labor Income of Immigrants in the US: A Special Focus on the Roles of Language and Home Country Income Level

Speaking the most-demanded languages is an asset in the labor market. However, coming from a high-income country may give immigrants an advantage in the labor market as those may have more transferable skills. This article investigates the determinants of the employability and labor income of immigrants and newcomer immigrants in the United States (US) labor market, specifically focusing on the role of language and income level of the home country. It applies the Heckman two-step selection procedure to the American Community Survey between 2000 and 2019. The findings show that immigrating from both high-income countries and countries where internationally most widely used languages (English, French, Portuguese, Spanish, and Chinese) are spoken gives immigrants an advantage in the US labor market compared to those from the countries where only national languages are spoken. This article emphasizes the key role of the income level of the home country on immigrants' labor income in the US. It contributes to the literature by employing the interaction terms of being from the same-income-level countries and the same-languages-speaking countries.

Model selection for time series with nonlinear trend

A two-step model selection procedure is proposed for autoregressive and moving-average (ARMA) model class. It is an adaptive least absolute shrinkage and selection operator (adLASSO) type model selection method that simultaneously chooses both the orders and significant lagged variables when the autoregressive and moving-average (ARMA) time series is contaminated with a nonlinear trend. The adLASSO is applied not to the observations, but to the detrended residuals. The proposed two-step adLASSO model selection procedure under some conditions can identify the true model with probability approaching one as the sample size increases, and the asymptotic properties of estimators are not affected by the replacement of observations with detrended residuals. The simulation studies show that the proposed method performs well with sample size as small as 50. The application of the method is illustrated by the annual varve thickness data collected from a location in Massachusetts.

Building a Machine-Learning Framework to Remotely Assess Parkinson's Disease Using Smartphones.

Parkinson's disease (PD) is a neurodegenerative disorder that affects multiple neurological systems. Traditional PD assessment is conducted by a physician during infrequent clinic visits. Using smartphones, remote patient monitoring has the potential to obtain objective behavioral data semi-continuously, track disease fluctuations, and avoid rater dependency. Smartphones collect sensor data during various active tests and passive monitoring, including balance (postural instability), dexterity (skill in performing tasks using hands), gait (the pattern of walking), tremor (involuntary muscle contraction and relaxation), and voice. Some of the features extracted from smartphone data are potentially associated with specific PD symptoms identified by physicians. To leverage large-scale cross-modality smartphone features, we propose a machine-learning framework for performing automated disease assessment. The framework consists of a two-step feature selection procedure and a generic model based on the elastic-net regularization. Using this framework, we map the PD-specific architecture of behaviors using data obtained from both PD participants and healthy controls (HCs). Utilizing these atlases of features, the framework shows promises to (a) discriminate PD participants from HCs, and (b) estimate the disease severity of individuals with PD. Data analysis results from 437 behavioral features obtained from 72 subjects (37 PD and 35 HC) sampled from 17 separate days during a period of up to six months suggest that this framework is potentially useful for the analysis of remotely collected smartphone sensor data in individuals with PD.

ALVEN: Algebraic learning via elastic net for static and dynamic nonlinear model identification

An algorithm is proposed that combines nonlinear feature generation and sparse regression to learn interpretable nonlinear models from noisy and limited data. This Algebraic Learning Via Elastic Net for Static and Dynamic Nonlinear Model Identification algorithm employs automated feature generation including families of ubiquitous chemical and biological nonlinear transformations. ALVEN balances model complexity and prediction accuracy through a two-step feature selection procedure, to produce an interpretable model useful for process applications while avoiding overfitting. The generalization to nonlinear dynamical systems, Dynamic ALVEN, is then described. The model accuracy of the algorithms is compared to well-established machine learning methods for a 3D printer and a chemical reactor.

Predictive mapping of bryophyte richness patterns in boreal forests using species distribution models and remote sensing data

Bryophytes represent an essential component of global biodiversity and play a significant role in many ecosystems, including boreal forests. In Canadian boreal forests, industrial exploitation of natural resources threatens bryophyte species and the ecological processes and services they support. However, the consideration of bryophytes in conservation issues is limited by current knowledge gaps on their distribution and diversity patterns. This is mainly due to the ineffectiveness of traditional field surveys to acquire information over large areas. Using remote sensing data in combination with species distribution models (SDMs), we aim to predict and map diversity patterns (in terms of richness) of i) total bryophytes, and ii) bryophyte guilds (mosses, liverworts and sphagna) in 28,436 km2 of boreal forests of Quebec (Canada). A bryophyte presence/absence database was used to develop four response variables: total bryophyte richness, moss richness, liverwort richness and sphagna richness. We pre-selected a group of 38 environmental predictors including climate, topography, soil moisture and drainage as well as vegetation. Then a final set of predictors was selected individually for each response variable through a two-step selection procedure. The Random Forest (RF) algorithm was used to develop spatially explicit regression models and to generate predictive cartography at 30 m resolution for the study area. Predictive mapping-associated uncertainty statistics were provided. Our models explained a significant fraction of the variation in total bryophyte and guild level richness, both in the calibration (42 to 52%) and validation sets (38 to 48%), outperforming models from previous studies. Vegetation (mainly NDVI) and climatic variables (temperature, precipitation, and freeze–thaw events) consistently appeared among the most important predictors for all bryophyte groups modeled. However, guild-level models identified differences in important factors determining the richness of each of the guilds and, therefore, in their predicted richness patterns. For example, the predictor number of days > 30 °C was especially relevant for liverworts, while drainage class, topographic position index and PALSAR HH-polarized L-band were identified among the most important predictors for sphagna. These differences have important implications for management and conservation strategies for bryophytes. This study provides evidence of the potential of remote sensing for assessing and making predictions on bryophyte diversity across the landscape.

Entrofy your cohort: A transparent method for diverse cohort selection.

Selecting a cohort from a set of candidates is a common task within and beyond academia. Admitting students, awarding grants, and choosing speakers for a conference are situations where human biases may affect the selection of any particular candidate, and, thereby the composition of the final cohort. In this paper, we propose a new algorithm, entrofy, designed to be part of a human-in-the-loop decision making strategy aimed at making cohort selection as just, transparent, and accountable as possible. We suggest embedding entrofy in a two-step selection procedure. During a merit review, the committee selects all applicants, submissions, or other entities that meet their merit-based criteria. This often yields a cohort larger than the admissible number. In the second stage, the target cohort can be chosen from this meritorious pool via a new algorithm and software tool called entrofy. entrofy optimizes differences across an assignable set of categories selected by the human committee. Criteria could include academic discipline, home country, experience with certain technologies, or other quantifiable characteristics. The entrofy algorithm then yields the approximation of pre-defined target proportions for each category by solving the tie-breaking problem with provable performance guarantees. We show how entrofy selects cohorts according to pre-determined characteristics in simulated sets of applications and demonstrate its use in a case study of Astro Hack Week. This two stage candidate and cohort selection process allows human judgment and debate to guide the assessment of candidates’ merit in step 1. Then the human committee defines relevant diversity criteria which will be used as computational parameters in entrofy. Once the parameters are defined, the set of candidates who meet the minimum threshold for merit are passed through the entrofy cohort selection procedure in step 2 which yields a cohort of a composition as close as possible to the computational parameters defined by the committee. This process has the benefit of separating the meritorious assessment of candidates from certain elements of their diversity and from some considerations around cohort composition. It also increases the transparency and auditability of the process, which enables, but does not guarantee, fairness. Splitting merit and diversity considerations into their own assessment stages makes it easier to explain why a given candidate was selected or rejected, though it does not eliminate the possibility of objectionable bias.

Computational Method for Identifying Malonylation Sites by Using Random Forest Algorithm.

As a newly uncovered post-translational modification on the ε-amino group of lysine residue, protein malonylation was found to be involved in metabolic pathways and certain diseases. Apart from experimental approaches, several computational methods based on machine learning algorithms were recently proposed to predict malonylation sites. However, previous methods failed to address imbalanced data sizes between positive and negative samples. In this study, we identified the significant features of malonylation sites in a novel computational method which applied machine learning algorithms and balanced data sizes by applying synthetic minority over-sampling technique. Four types of features, namely, amino acid (AA) composition, position-specific scoring matrix (PSSM), AA factor, and disorder were used to encode residues in protein segments. Then, a two-step feature selection procedure including maximum relevance minimum redundancy and incremental feature selection, together with random forest algorithm, was performed on the constructed hybrid feature vector. An optimal classifier was built from the optimal feature subset, which featured an F1-measure of 0.356. Feature analysis was performed on several selected important features. Results showed that certain types of PSSM and disorder features may be closely associated with malonylation of lysine residues. Our study contributes to the development of computational approaches for predicting malonyllysine and provides insights into molecular mechanism of malonylation.

Industrial equipment reliability estimation: A Bayesian Weibull regression model with covariate selection

A three-state continuous-time semi-Markov process is used to model the degradation of an industrial equipment. The transition times are assumed Weibull-distributed and influenced by a set of covariates. A Weibull Regression Model is developed within the Bayesian probability framework, to account for the influence of these covariates and estimate the model parameters with the related uncertainty, on the basis of few data and expert judgment. The number of covariates is reduced by a two-step selection procedure derived from the condition monitoring engineering practice. The developed model enables estimating reliability and time-dependent state probabilities for a component degrading in given operational and ambient conditions, represented by a vector of covariates. The model is illustrated by way of a real case study concerning the degradation process affecting diaphragm valves used in the biopharmaceutical industry.

A novel two-step pseudo-response based adaptive harmonic balance method for dynamic analysis of nonlinear structures

Harmonic balance method (HBM) is one of the most popular and powerful methods, which is used to obtain response of nonlinear vibratory systems in frequency domain. The main idea of the method is to express the response of the system in Fourier series and converting the nonlinear differential equations of motion into a set of nonlinear algebraic equations. System response can be obtained by solving this nonlinear equation set in terms of the unknown Fourier coefficients. The accuracy of the solution is greatly affected by the number of harmonics included in the method and it is enhanced as the number of harmonics increases at the expense of computational time; hence, advantage of HBM over time integration method is lost. Therefore, it is desirable to use an adaptive algorithm where the number of harmonics can be optimized in terms of both accuracy and computational effort. In this paper a new adaptive harmonic balance method (AHBM) for the dynamic analysis of nonlinear structures is developed. The new method employs a two-step harmonic selection procedure where the criteria used are based on simple magnitude comparisons that make it easy to understand and program the method. A novel pseudo-response calculation method, which is used at the second harmonic selection step, is developed in order to estimate the response of the nonlinear system with, approximately, no additional computational cost. Due to the two-step harmonic selection procedure, the method eliminates unnecessary harmonics in the response calculation; hence, it is capable of increasing the computational efficiency of HBM significantly. Several case studies are given in order to show the applicability of the proposed adaptive harmonic balance method.

Evaluating seasonal effect on forest leaf area index mapping using multi-seasonal high resolution satellite pléiades imagery

The forest canopy leaf area index (LAI) is an important structural variable directly affecting functions and structures of terrestrial plant ecosystems. Optical remote sensing techniques may provide an alternative in estimating and mapping plant LAI. However, existing studies on using very high resolution (VHR) multitemporal satellite imagery to map and investigate the seasonal effect on plant LAI at a landscape scale are rare. In this study, we proposed to map and analyze forest LAI using four seasonal Pléiades images and corresponding in situ seasonal LAI measurements collected over a natural forest area in the City of Tampa, Florida, USA. A subset of selected spectral/textural features was used to develop pixel-based seasonal LAI regression models through a two-step feature selection procedure and a canonical correlation analysis. Finally, seasonal changes of the mapped LAIs were analyzed and assessed. Several interesting experimental results were created through this study, including: (i) a set of optimal texture parameters for extracting the 1st- and 2nd-order grey level statistical textures from the Pléiades imagery was determined as a window size 5 × 5, a direction 90° and pixel displacement 4 pixels; (ii) textural features were more important than spectral features in estimating and mapping forest LAI, and red band has a higher power in mapping forest LAI than other three multispectral bands; (iii) the late spring Pleiades image resulted in the highest accuracy for estimating and mapping forest LAI; and (iv) there exists a significant seasonal change of forest LAI in the study area and the seasonal effect on forest LAI mapping can be assessed by using the multi-seasonal VHR satellite imagery at a landscape scale. A novel significance for this study is that it is the first time using both spectral and textural information extracted from the multi-seasonal VHR satellite images to assess the seasonal effect on forest LAI mapping at a landscape scale. Since the experimental results and findings were derived from a relatively small study area, further testing and validation work is needed over different forest ecosystems at a landscape scale.

Bachelor completion and dropout rates of selected, rejected and lottery-admitted medical students in the Netherlands

BackgroundEvidence for the effectiveness of the selection of medical students is weak. This study aimed to examine the added value of a two-step selection procedure (first step non-academic, second step academic tests) to a pre-university GPA-based lottery procedure. Because previous research has suggested that participation in selection (regardless of the outcome) is a predictor of study success, this study is the first to include students who initially applied for selection, then refrained from (actively) participating in selection and were eventually admitted through lottery.MethodsBachelor completion and dropout rates of selected (n = 416) and lottery-admitted students from four cohorts (2006–2009) were compared using logistic regression analysis. Four groups of lottery-admitted students were distinguished: students who were rejected after step 2 (n = 57), were rejected after step 1 (n = 169), withdrew during selection step 1 (n = 42) and students who only applied for lottery (n = 366). Covariates included gender, age, pre-university GPA and cohort.ResultsThere was a significant association between admission group and obtaining a bachelor degree in three years. Selected students were more likely to obtain a bachelor degree within three years (64.2% versus 51.6%; OR = 1.7) or four years (81.5% versus 74.3%; OR = 1.6) than students who only applied to a lottery (p < 0.05); selected students also seemed more likely to obtain all Year-1 course credits than students who withdrew during step 1 (40.4% versus 21.4%; OR = 2.3; p < 0.05). We found no significant association between dropout and admission groups. Students rejected at step 1 or 2 did not perform significantly different from selected students on any of the outcome measures.ConclusionsThe findings indicated that students at risk for study delay in the preclinical phase in our context were more likely to refrain from applying to a demanding selection procedure when a less demanding alternative was available. We found no significant associations between the non-academic and academic selection steps and bachelor completion and dropout rates. These findings suggest that the presence of the selection was more important than these specific selection components. In follow-up research, we plan to investigate the associations between the admission groups and outcome measures in the clinical phase.

Penalty functions and two-step selection procedure based DIRECT-type algorithm for constrained global optimization

Applied optimization problems often include constraints. Although the well-known derivative-free global-search DIRECT algorithm performs well solving box-constrained global optimization problems, it does not naturally address constraints. In this article, we develop a new algorithm DIRECT-GLce for general constrained global optimization problems incorporating two-step selection procedure and penalty function approach in our recent DIRECT-GL algorithm. The proposed algorithm effectively explores hyper-rectangles with infeasible centers which are close to boundaries of feasibility and may cover feasible regions. An extensive experimental investigation revealed the potential of the proposed approach compared with other existing DIRECT-type algorithms for constrained global optimization problems, including important engineering problems.

AtbPpred: A Robust Sequence-Based Prediction of Anti-Tubercular Peptides Using Extremely Randomized Trees

Mycobacterium tuberculosis is one of the most dangerous pathogens in humans. It acts as an etiological agent of tuberculosis (TB), infecting almost one-third of the world's population. Owing to the high incidence of multidrug-resistant TB and extensively drug-resistant TB, there is an urgent need for novel and effective alternative therapies. Peptide-based therapy has several advantages, such as diverse mechanisms of action, low immunogenicity, and selective affinity to bacterial cell envelopes. However, the identification of anti-tubercular peptides (AtbPs) via experimentation is laborious and expensive; hence, the development of an efficient computational method is necessary for the prediction of AtbPs prior to both in vitro and in vivo experiments. To this end, we developed a two-layer machine learning (ML)-based predictor called AtbPpred for the identification of AtbPs. In the first layer, we applied a two-step feature selection procedure and identified the optimal feature set individually for nine different feature encodings, whose corresponding models were developed using extremely randomized tree (ERT). In the second-layer, the predicted probability of AtbPs from the above nine models were considered as input features to ERT and developed the final predictor. AtbPpred respectively achieved average accuracies of 88.3% and 87.3% during cross-validation and an independent evaluation, which were ~8.7% and 10.0% higher than the state-of-the-art method. Furthermore, we established a user-friendly webserver which is currently available at http://thegleelab.org/AtbPpred. We anticipate that this predictor could be useful in the high-throughput prediction of AtbPs and also provide mechanistic insights into its functions.

Forecasting bank failures and stress testing: A machine learning approach

This paper presents a forecasting model of bank failures based on machine-learning. The proposed methodology defines a linear decision boundary that separates the solvent banks from those that failed. This setup generates a novel alternative stress-testing tool. Our sample of 1443 U.S. banks includes all 481 banks that failed during the period 2007–2013. The set of explanatory variables is selected using a two-step feature selection procedure. The selected variables were then fed to a support vector machines forecasting model, through a training–testing learning process. The model exhibits a 99.22% overall forecasting accuracy and outperforms the well-established Ohlson’s score.

IProt-Sub: a comprehensive package for accurately mapping and predicting protease-specific substrates and cleavage sites.

Regulation of proteolysis plays a critical role in a myriad of important cellular processes. The key to better understanding the mechanisms that control this process is to identify the specific substrates that each protease targets. To address this, we have developed iProt-Sub, a powerful bioinformatics tool for the accurate prediction of protease-specific substrates and their cleavage sites. Importantly, iProt-Sub represents a significantly advanced version of its successful predecessor, PROSPER. It provides optimized cleavage site prediction models with better prediction performance and coverage for more species-specific proteases (4 major protease families and 38 different proteases). iProt-Sub integrates heterogeneous sequence and structural features and uses a two-step feature selection procedure to further remove redundant and irrelevant features in an effort to improve the cleavage site prediction accuracy. Features used by iProt-Sub are encoded by 11 different sequence encoding schemes, including local amino acid sequence profile, secondary structure, solvent accessibility and native disorder, which will allow a more accurate representation of the protease specificity of approximately 38 proteases and training of the prediction models. Benchmarking experiments using cross-validation and independent tests showed that iProt-Sub is able to achieve a better performance than several existing generic tools. We anticipate that iProt-Sub will be a powerful tool for proteome-wide prediction of protease-specific substrates and their cleavage sites, and will facilitate hypothesis-driven functional interrogation of protease-specific substrate cleavage and proteolytic events.

Forecasting Bank Failures and Stress Testing: A Machine Learning Approach

In this paper, we present a forecasting model of bank failures based on machine-learning. The proposed methodology defines a linear decision boundary separating the solvent from the failed banks. This setup generates a novel alternative stress testing tool. Our sample of 1443 U.S. banks includes all 481 failed banks during the 2007-2013 period. The set of explanatory variables is selected using a two-step feature selection procedure. The selected variables were then fed to a to Support Vector Machines forecasting model, through a training-testing learning process. The model exhibits 99.22% overall forecasting accuracy and outperforms the well-established Ohlson’s score.