Model Selection Research Articles

Exploring the power of data-driven models for groundwater system conceptualization: a case study of the Grazer Feld Aquifer, Austria

AbstractDeveloping a reliable conceptual model is crucial for analyzing groundwater systems. An essential part of the aquifer conceptualization is the identification of the hydrological stresses that control the hydraulic head fluctuations. By effectively capturing and understanding these stresses, the propagation of potential errors and uncertainties through subsequent modeling steps can be minimized. This study aims to test data-driven models as screening models for conceptualizing a groundwater system. The case study is applied to the Grazer Feld Aquifer in southeast Austria. Time series models are applied to: (1) identify the stresses likely influencing the observed head fluctuations and their spatial variability; (2) identify locations where a lack of understanding of head fluctuations exists; and (3) discuss the limitations and opportunities associated with data-driven models to support system conceptualization. Time series models were created for 144 monitoring wells where sufficient head observations were available during the calibration period (2005–2015). A total of 576 models were developed, incorporating the combinations of stresses: recharge, river level, and a step trend. Following the model selection process, each model was categorized based on its performance and divided into four groups. At 88 sites, recharge and river level variations were identified as the primary controlling stresses influencing head fluctuations. The inclusion of the step trend was found to be necessary at five sites to accurately simulate heads due to dam construction. The application of data-driven models in this study enhanced the identification of key aquifer stresses, facilitating a more informed understanding of the groundwater system.

Covariance Assisted Multivariate Penalized Additive Regression (CoMPAdRe)

We propose a new method for the simultaneous selection and estimation of multivariate sparse additive models with correlated errors. Our method called Covariance Assisted Multivariate Penalized Additive Regression (CoMPAdRe) simultaneously selects among null, linear, and smooth non-linear effects for each predictor while incorporating joint estimation of the sparse residual structure among responses, with the motivation that accounting for inter-response correlation structure can lead to improved accuracy in variable selection and estimation efficiency. CoMPAdRe is constructed in a computationally efficient way that allows the selection and estimation of linear and non-linear covariates to be conducted in parallel across responses. Compared to single-response approaches that marginally select linear and non-linear covariate effects, we demonstrate in simulation studies that the joint multivariate modeling leads to gains in both estimation efficiency and selection accuracy, of greater magnitude in settings where signal is moderate relative to the level of noise. We apply our approach to protein-mRNA expression levels from multiple breast cancer pathways obtained from The Cancer Proteome Atlas and characterize both mRNA-protein associations and protein-protein subnetworks for each pathway. We find non-linear mRNA-protein associations for the Core Reactive, EMT, PIK-AKT, and RTK pathways. Supplementary Materials are available online.

Complacency in the quantification and reporting of climate impacts as carbon dioxide equivalent emissions

PurposeThe choice of characterization model has the potential to profoundly influence LCA results and conclusions. It is therefore a requirement of ISO 14044:2006 that the selection of characterization model shall be both justified and consistent with goal and scope. The purpose of this article was to examine current practices regarding the characterization of GHG emissions and reporting as CO2 equivalent emissions.MethodsA survey of practice was conducted across articles recently published in the International Journal of Life Cycle Assessment. Each article was examined using seven predetermined questions covering reporting of CO2 equivalent emissions and justification of climate metric used, interpretation of results in relation to climate measures or goals, and the implications of choosing an alternative climate metric.Results and discussionOf the sample of 85 articles, more than half reported GHG emissions as CO2 equivalent emissions. Of these, 81% unambiguously reported the climate metric used. Most often, there was no justification for the choice of a characterization model. Where a justification was given, the most common reason was alignment with the requirements of a program or PCR document. In some cases, the choice of characterization model was determined by the choice of impact assessment method (e.g., CML, ReCiPe). In other cases, the choice of characterization model was based on the desire to compare results to other studies. It is noted that none of the abovementioned reasons is scientific justification related to a stated climate objective. Not surprisingly, most studies made no attempt to interpret results reported as CO2 equivalent emissions in relation to climate measures or goals, and most did not discuss the potential implications of alternative climate metrics. For almost half of the articles, the choice of climate metric was assessed as potentially having major implications for decision-making or comparison to alternative systems.ConclusionsBased on this survey, it is evident that key aspects of ISO 14044:2006 are routinely not being followed. When GHG emissions are aggregated into a single impact category indicator result, there is a loss of transparency about climate impacts over time and the potential to unknowingly trade short-term climate benefits against the exacerbation of the difficulties of achieving long-term climate stabilization. As such, whenever GHG emissions are reported as CO2 equivalent emissions, it is imperative that the choice of characterization model is unambiguously reported and justified, that results are interpreted in relation to environmental outcomes, and that the potential implications of selecting alternative models are discussed.Graphical

Novel two-parameter quadratic exponential distribution: Properties, simulation, and applications

This research paper introduces a novel two-parameter quadratic exponential distribution (NTPQED), thoroughly examining its statistical properties and practical applications. The study delves into essential characteristics of the distribution, including its asymptotic behavior, moments, order statistics, and entropies. Additionally, we present fuzzy reliability, value at risk, mean excess function, limited expected value function, tail value at risk, and tail variance. A comprehensive simulation study evaluates fifteen parameter estimation methods. To illustrate the practical relevance of the proposed distribution, three sets of real-world data are analyzed. The performance of the proposed distribution model is rigorously assessed using various model selection criteria and goodness-of-fit test statistics. Empirical findings consistently demonstrate the proposed distribution's superiority over existing models, highlighting its potential for multiple applications.

Influence of device configuration and noise on a machine learning predictor for the selection of nanoparticle small-angle X-ray scattering models.

Small-angle X-ray scattering (SAXS) is a widely used method for nanoparticle characterization. A common approach to analysing nanoparticles in solution by SAXS involves fitting the curve using a parametric model that relates real-space parameters, such as nanoparticle size and electron density, to intensity values in reciprocal space. Selecting the optimal model is a crucial step in terms of analysis quality and can be time-consuming and complex. Several studies have proposed effective methods, based on machine learning, to automate the model selection step. Deploying these methods in software intended for both researchers and industry raises several issues. The diversity of SAXS instrumentation requires assessment of the robustness of these methods on data from various machine configurations, involving significant variations in the q-space ranges and highly variable signal-to-noise ratios (SNR) from one data set to another. In the case of laboratory instrumentation, data acquisition can be time-consuming and there is no universal criterion for defining an optimal acquisition time. This paper presents an approach that revisits the nanoparticle model selection method proposed by Monge et al. [Acta Cryst. (2024), A80, 202-212], evaluating and enhancing its robustness on data from device configurations not seen during training, by expanding the data set used for training. The influence of SNR on predictor robustness is then assessed, improved, and used to propose a stopping criterion for optimizing the trade-off between exposure time and data quality.

Analyzing the critical steps in deep learning-based stock forecasting: a literature review

Stock market or individual stock forecasting poses a significant challenge due to the influence of uncertainty and dynamic conditions in financial markets. Traditional methods, such as fundamental and technical analysis, have been limited in coping with uncertainty. In recent years, this has led to a growing interest in using deep learning-based models for stock prediction. However, the accuracy and reliability of these models depend on correctly implementing a series of critical steps. These steps include data collection and analysis, feature extraction and selection, noise elimination, model selection and architecture determination, choice of training-test approach, and performance evaluation. This study systematically examined deep learning-based stock forecasting models in the literature, investigating the effects of these steps on the model’s forecasting performance. This review focused on the studies between 2020–2024, identifying influential studies by conducting a systematic literature search across three different databases. The identified studies regarding seven critical steps essential for creating successful and reliable prediction models were thoroughly examined. The findings from these examinations were summarized in tables, and the gaps in the literature were detailed. This systematic review not only provides a comprehensive understanding of current studies but also serves as a guide for future research.

Pengaruh Komisaris Independen, Capital Intensity Dan Financial Distress Terhadap Konservatisme Akuntansi

This research aims to test and prove the influence of independent commissioners, capital intensity and financial distress on accounting conservatism” (Empirical Study on LQ 45 index companies listed on the BEI for the 2018-2023 period).The type of research used is a quantitative method. The type of data used is secondary data in the form of financial reports of LQ 45 companies listed on the Indonesia Stock Exchange for the period 2018-2023. Samples were collected using the purposive sampling method. The number of companies that met the research sample criteria was 15 companies out of 45 companies with a research period of 6 years of observation, so that a total of 90 research samples were obtained. Data processing using the Eviews 10 Statistical Program. Next, carrying out various tests including descriptive statistical tests, panel data model tests, model selection tests, classical assumption tests and hypothesis tests. So the research results prove that independent commissioners, capital intensity and financial distress simultaneously influence accounting conservatism. Independent commissioners partially influence accounting conservatism. capital intensity influences accounting conservatism. Financial distress has no effect on accounting conservatism.

Integrative Modeling of Protein-Polypeptide Complexes by Bayesian Model Selection using AlphaFold and NMR Chemical Shift Perturbation Data.

Protein-polypeptide interactions, including those involving intrinsically-disordered peptides and intrinsically-disordered regions of protein binding partners, are crucial for many biological functions. However, experimental structure determination of protein-peptide complexes can be challenging. Computational methods, while promising, generally require experimental data for validation and refinement. Here we present CSP_Rank, an integrated modeling approach to determine the structures of protein-peptide complexes. This method combines AlphaFold2 (AF2) enhanced sampling methods with a Bayesian conformational selection process based on experimental Nuclear Magnetic Resonance (NMR) Chemical Shift Perturbation (CSP) data and AF2 confidence metrics. Using a curated dataset of 108 protein-peptide complexes from the Biological Magnetic Resonance Data Bank (BMRB), we observe that while AF2 typically yields models with excellent consistency with experimental CSP data, applying enhanced sampling followed by data-guided conformational selection routinely results in ensembles of structures with improved agreement with NMR observables. For two systems, we cross-validate the CSP-selected models using independently acquired nuclear Overhauser effect (NOE) NMR data and demonstrate how CSP and NMR can be combined using our Bayesian framework for model selection. CSP_Rank is a novel method for integrative modeling of protein-peptide complexes and has broad implications for studies of protein-peptide interactions and aiding in understanding their biological functions.

Automated model selection for multivariate anomaly detection in manufacturing systems

AbstractAs machine learning is widely applied to improve the efficiency and effectiveness of manufacturing systems, the automated selection of appropriate algorithms and hyperparameters becomes increasingly important. This paper presents a model selection approach to multivariate anomaly detection for applications in manufacturing systems using a multi-output regression-based meta-learning method. The proposed method exploits the capabilities of meta-learning to explore and learn the intricate relationships within multivariate data sets in order to select the best anomaly detection model. It also facilitates the construction of an ensemble of algorithms with dynamically assigned weights based on their respective performance levels. In addition to the framework, new meta-features for the application domain are presented and evaluated. Experiments show the proposed method can be successfully applied to achieve significantly better results than benchmark approaches. This enables an automated selection of algorithms that can be used for enhanced anomaly detection under changing operating conditions.

Diagnosis of ophthalmologic diseases in canines based on images using neural networks for image segmentation

The primary challenge in diagnosing ocular diseases in canines based on images lies in developing an accurate and reliable machine learning method capable of effectively segmenting and diagnosing these conditions through image analysis. Addressing this challenge, the study focuses on developing and rigorously evaluating a machine learning model for diagnosing ocular diseases in canines, employing the U-Net neural network architecture as a foundational element of this investigation.Through this extensive evaluation, the authors identified a model that exhibited good reliability, achieving prediction scores with an Intersection over Union (IoU) exceeding 80 %, as measured by the Jaccard index. The research methodology encompassed a systematic exploration of various neural network backbones (VGG, ResNet, Inception, EfficientNet) and the U-Net model, combined with an extensive model selection process and an in-depth analysis of a custom training dataset consisting of historical images of different medical symptoms and diseases in dog eyes.The results indicate a fairly high degree of accuracy in the segmentation and diagnosis of ocular diseases in canines, demonstrating the model's effectiveness in real-world applications. In conclusion, this potentially makes a significant contribution to the field by utilizing advanced machine-learning techniques to develop image-based diagnostic routines in veterinary ophthalmology. This model's successful development and validation offer a promising new tool for veterinarians and pet owners, enhancing early disease detection and improving health outcomes for canine patients.

An Evaluation of an Algorithm for the Selection of Flexible Survival Models for Cancer Immunotherapies: Pass or Fail?

An Evaluation of an Algorithm for the Selection of Flexible Survival Models for Cancer Immunotherapies: Pass or Fail?

Prediction of pathological response and lymph node metastasis after neoadjuvant therapy in rectal cancer through tumor and mesorectal MRI radiomic features

Establishing predictive models for the pathological response and lymph node metastasis in locally advanced rectal cancer (LARC) treated with neoadjuvant chemoradiotherapy (nCRT) based on MRI radiomic features derived from the tumor and mesorectal compartment (MC). This study included 209 patients with LARC who underwent rectal MRI both before and after nCRT. The patients were divided into a training set (n = 146) and a test set (n = 63). Regions of interest (ROIs) for the tumor and MC were delineated on both pre- and post-nCRT MRI images. Radiomic features were extracted, and delta radiomic features were computed. The predictive endpoints were pathological complete response (pCR), pathological good response (pGR), and lymph node metastasis (LNM). Feature selection for various models involved sequentially removing features with a correlation coefficient > 0.9, and features with P-values ≥ 0.05 in univariate analysis, followed by LASSO regression on the remaining features. Logistic regression models were developed, and their performance was evaluated using the area under the receiver operating characteristic curve (AUC). Among the 209 LARC patients, the number of patients achieving pCR, pGR, and LNM were 44, 118, and 40, respectively. The optimal model for predicting each endpoint is the combined model that incorporates pre- and delta-radiomics features for both the tumor and MC. These models exhibited superior performance with AUC values of 0.874 (for pCR), 0.801 (for pGR), and 0.826 (for LNM), outperforming the MRI tumor regression grade (mrTRG) which yielded AUC values of 0.800, 0.715, and 0.603, respectively. The results demonstrate the potential utility of the tumor and MC radiomics features, in predicting treatment efficacy among LARC patients undergoing nCRT.

Determining magnetic structures in GSAS-II using the Bilbao Crystallographic Server tool k-SUBGROUPSMAG.

The embedded call to a special version of the web-based Bilbao Crystallographic Server tool k-SUBGROUPSMAG from within GSAS-II to form a list of all possible commensurate magnetic subgroups of a parent magnetic grey group is described. It facilitates the selection and refinement of the best commensurate magnetic structure model by having all the analysis tools including Rietveld refinement in one place as part of GSAS-II. It also provides the chosen magnetic space group as one of the 1421 possible standard Belov-Neronova-Smirnova forms or equivalent non-standard versions.

Enhancing structural damage evaluation of PC girder bridges through digital twinning Bayesian model updating

Developing a unified analytical framework to assess the structural performance of deteriorated prestressed concrete (PC) girder bridges is important for their maintenance. This study aims to construct a physics-based digital twin framework for PC girder bridges incorporating Bayesian model selection and updating, a high-fidelity three-dimensional (3D) finite element (FE) model, and a method for simulating prestressed tendon damage. In this study, three PC girder specimens were designed and tested to evaluate the effects of tendon rupture near the anchorage region on structural performance. Bayesian model selection was used to determine the most plausible model class based on the estimated model evidence. The most probable values (MPVs) of the model parameters, derived from the estimated posterior probability density functions (PDFs), were then used to calibrate the FE model, which reflects a healthy state of the PC girder bridge. This updated FE model serves as the baseline for conducting what-if scenario simulations. Subsequently, nonlinear segments of the constitutive model, obtained from coupon tests, and a simulation method for tendon rupture were incorporated into the updated FE model. Nonlinear static simulation results obtained through the updated model were found to be consistent with both the observed crack pattern and the load–deflection curve. The reliability of the digital twinning FE model for assessing tendon damage and for predicting the residual load-bearing capacity was verified through comparison with measured data. Therefore, the digital twinning Bayesian model updating approach shows potential applications in continuous structural health monitoring (SHM) of PC bridges.

A VAR Model for Non-Performing-Loans in Albania

The objective of this paper is to distinguish the relationship of macroeconomic and macro financial factors with non-performing loans in the Albanian banking system for the last 10 years, through a VAR model. VAR models are used to discover the relationship between variables, while the model selection criteria are used to select the appropriate lag. The chosen lag is 4 and the macroeconomic variables besides NPL included in the study are unemployment, economic growth and consumer price index. The results of the model proved that NPL ratio is sensitive to previous periods of NPL movement. Other macroeconomic variables are insignificant related to actual NPL. Furthermore, through the representation of fan charts we expect a reduction on NPL ratio on the upcoming quarter, which signifies a stronger economy.

LDAPrototype: a model selection algorithm to improve reliability of latent Dirichlet allocation

Latent Dirichlet allocation (LDA) is a popular method for analyzing large text corpora, but it suffers from instability due to its reliance on random initialization. This results in different outcomes for replicated runs, hindering reproducibility. To address this, we introduce LDAPrototype, a new approach for selecting the most representative LDA run from multiple replications on the same dataset. LDAPrototype enhances the reliability of LDA conclusions by ensuring greater similarity between replications compared to traditional LDA runs or models chosen based on perplexity or NPMI. A key feature of LDAPrototype is its use of a novel model similarity measure called S-CLOP (Similarity of multiple sets by Clustering with LOcal Pruning). It is based on topic similarities, for which we compare the usage of measures like the thresholded Jaccard coefficient, cosine similarity, Jensen-Shannon divergence, and rank-biased overlap. The effectiveness of LDAPrototype is demonstrated through its application to six real datasets, including newspaper articles and tweets. The results show improved reproducibility and reliability in topic modeling outcomes. LDAPrototype’s approach is noteworthy for its practical applicability, comprehensibility, ease of implementation, and computational efficiency. Furthermore, the algorithm’s concept can be generalized to other topic modeling procedures that characterize topics through word distributions, making it a versatile tool in text data analysis.

An intriguing way to synthesize a TiO2–ZnO hybrid nanostructure for the pragmatic application as a photocatalyst

In an attempt to counter the problem of treating wastewater more efficiently, we report the study on photocatalytic properties of metal oxide nanostructures under different light sources. A novel sonication-assisted refluxing method was employed to synthesize a binary heterostructure consisting of TiO2 and ZnO. XRD and EDS analysis of synthesized materials confirm the presence of both TiO2 and ZnO. Also, the change in the intensity of XRD peaks and elemental composition are well observed with a change in molar concentration of Ti and Zn. HR-TEM micrographs show nanostructures having hybrid Janus and Core-shell type structures. The XPS spectra were probed to identify the surface chemical species associated with the prepared heterostructures. Raman spectra confirm the smaller ZnO particles attached over agglomerated TiO2 as a shell layer as depicted in TEM results. Optical studies exhibit the modification in the band structure of the materials due to the formation of the heterostructure. The synthesized hybrid catalyst shows impressive results in photocatalytic performance. The catalytic potency of the prepared heterostructures under UV light is reminiscent of Anatase while it gets convalescent with an increment of Zn content under visible light. The photocatalytic performance of synthesized materials under sunlight shows their potential to be utilized as commercial catalysts for direct application in wastewater treatment. The kinetic study of the degradation data reveals interesting facets of the catalysis. It suggests that the photo-corrosion of the catalyst inordinately changes the rate of photocatalytic reaction which ultimately affects the photocatalytic efficiency of the catalyst. The study also discusses the role of kinetic and validation parameters in the selection of the best-fitted kinetic model with realistic arguments.

Survey of federated learning in intrusion detection

Intrusion detection methods are crucial means to mitigate network security issues. However, the challenges posed by large-scale complex network environments include local information islands, regional privacy leaks, communication burdens, difficulties in handling heterogeneous data, and storage resource bottlenecks. Federated learning has the potential to address these challenges by leveraging widely distributed and heterogeneous data, achieving load balancing of storage and computing resources across multiple nodes, and reducing the risks of privacy leaks and bandwidth resource demands. This paper reviews the process of constructing federated learning based intrusion detection system from the perspective of intrusion detection. Specifically, it outlines six main aspects: application scenario analysis, federated learning methods, privacy and security protection, selection of classification models, data sources and client data distribution, and evaluation metrics, establishing them as key research content. Subsequently, six research topics are extracted based on these aspects. These topics include expanding application scenarios, enhancing aggregation algorithm, enhancing security, enhancing classification models, personalizing model and utilizing unlabeled data. Furthermore, the paper delves into research content related to each of these topics through in-depth investigation and analysis. Finally, the paper discusses the current challenges faced by research, and suggests promising directions for future exploration.

Challenging the Atacama desert: Agronomic and water conditions for pre-Hispanic maize agriculture in hyper arid environments inferred by δ18O isotopes

During the Formative Period (ca.2400–950 years BP), pre-Hispanic farmers in Pampa del Tamarugal developed a complex agricultural system in the hyper-arid Atacama Desert in Tarapacá, northern Chile. This system involved numerous agricultural fields congregated near the Tarapacá Valley’s perennial stream and the Guatacondo ravine’s ephemeral stream. Well-established villages such as Caserones, Pircas, Ramaditas, and Guatacondo accompanied these developments. However, the importance of understanding the water sources has been neglected despite the efforts to understand the relationships between pre-Hispanic agriculture and their living environment. Thus, we presumed the use of local water sources without a clear understanding of their exploitation methods or the associated technological implications. Furthermore, there is limited research on groundwater use in Tarapacá. This research aims to understand water sources used in the cultivation of maize (Zea mays) using δ18O isotope values obtained from pre-Hispanic maize kernels in Tarapacá. We compared these values with published δ18O values of water sources and applied a generalized linear model (GLM) with a Gaussian distribution, performing a Tukey’s post hoc test for multiple comparisons of means with heteroscedasticity-consistent covariance estimation. The best-fit model was identified using a stepwise model selection procedure based on the Bayesian information criterion (BIC). Our results indicate that mean δ18O values of organic matter from maize kernels range from 24.73 ‰ to 31.65 ‰. The best performing model on δ18O only included Group (BIC=298.2) as the explanatory variable as Period, Site, and Weight had no effects. These findings point towards a significant statistical relationship between the δ18O values of organic matter derived from maize kernels and the specific geographic regions they originate from. These values also show an enrichment of δ18O isotope in Tarapacá samples, except for Pica 8. These results indicate diverse agricultural strategies that utilized different water sources including perennial flow in the Tarapacá River, ephemeral runoff at Guatacondo, and groundwater in the Pampa del Tamarugal. Also, due to the enrichment of δ18O, we suggest that the circular structures found among the fields were used as water reservoirs leading to the observed enrichment of δ18O isotope values. We discuss the role of ancient agriculture technology in water management, the role of maize, cultural strategies, and the relationships with their environment. In conclusion, pre-Hispanic farmers managed limited water sources successfully despite intermittent drought for hundreds of years until the 1970 s, when agriculture was abandoned due to multiple factors.

FRACTAL-INSPIRED SENTIMENT ANALYSIS: EVALUATION OF LARGE LANGUAGE MODELS AND DEEP LEARNING METHODS

Sentiment analysis is a vital task in natural language processing (NLP) that aims to identify and extract the emotional states and opinions of text. In this study, we conduct a comprehensive comparison of large language models (LLMs), such as ChatGPT and Google Bard, with conventional methods in sentiment analysis. We employ a rigorous evaluation framework that covers four essential metrics: accuracy, precision, recall, and the [Formula: see text]1-score. Our results reveal that TextBlob outperforms other methods, achieving an impressive accuracy of 69% and precision of 83%. On the other hand, Bard shows a relatively poor performance, with only 39% accuracy and 46% precision. This study offers valuable insights into the diverse capabilities of AI models in sentiment analysis. A key finding of this study is the importance of model selection according to the specific requirements of the task. Each model has its own strengths and weaknesses, which are reflected in their performance profiles. Moreover, the context in which these models operate is crucial. For instance, ChatGPT generates varied responses, Bard struggles with multiple sentences, and Robustly Optimized BERT Pretraining Approach (RoBERTa) balances precision and recall. This study also reveals the performance gap between LLMs and state-of-the-art deep learning methods. We believe this work will inspire future research and applications of ChatGPT and similar AI models in sentiment analysis and related tasks.