Multiple Methods Research Articles

Classification and Explanation of Iron Deficiency Anemia from Complete Blood Count Data Using Machine Learning

Background: Currently, discriminating Iron Deficiency Anemia (IDA) from other anemia requires an expensive test (serum ferritin). Complete Blood Count (CBC) tests are less costly and more widely available. Machine learning models have not yet been applied to discriminating IDA but do well for similar tasks. Methods: We constructed multiple machine learning methods to classify IDA from CBC data using a US NHANES dataset of over 19,000 instances, calculating accuracy, precision, recall, and precision AUC (PR AUC). We validated the results using an unseen dataset from Kenya, using the same model. We calculated ranked feature importance to explain the global behavior of the model. Results: Our model classifies IDA with a PR AUC of 0.87 and recall/sensitivity of 0.98 and 0.89 for the original dataset and an unseen Kenya dataset, respectively. The explanations indicate that low blood level of hemoglobin, higher age, and higher Red Blood Cell distribution width were most critical. We also found that optimization made only minor changes to the explanations and that the features used remained consistent with professional practice. Conclusions: The overall high performance and consistency of the results suggest that the approach would be acceptable to health professionals and would support enhancements to current automated CBC analyzers.

Summary of the clinical characteristics of tuberculous otitis media

Objective:The purpose of this study was to analyze and summarize the clinical characteristics and diagnostic methods of tuberculous otitis media（TOM）, to enrich clinical experience in diagnosis and treatment of tuberculous otitis media, so as to reduce missed diagnosis and misdiagnosis, and facilitate timely and effective therapy for better prognosis. Methods:This study retrospectively analyzed the clinical data of patients with tuberculous otitis media who were hospitalized in the Ear ward of our hospital and received surgical treatment from 2008 to 2022. The data of patients' clinical characteristics, radiological examination, intraoperative findings and therapeutic strategies were recorded and summarized. Results:A total of 23 cases （26 ears） of tuberculous otitis media were included in this retrospective study. The most common clinical symptoms were otorrhea（thin odorless fluid）（100%） and conductive hearing loss（100%）, with a high incidence of facial paralysis（23.1%）. It was not sensitive to traditional antibiotic treatment, eg. Levofloxacin （50% effective rate only）, and relapsed soon after drug withdrawal. It was revealed that all the surgical views had gray and white tough granulation tissue hyperplasia（100%）, and 23.1% with caseous necrosis. The purpose of surgery was to clear the lesion, reduce the recurrence rate of suppurative infection, and repair the function （hearing reconstruction or facial nerve decompression） as appropriate. The paraffin pathology of granulation tissue were reported as typical granulomatous inflammation and caseous necrosis with positive acid-fast staining, which was consistent with tuberculosis. Conclusion:It was easily confused by the clinical manifestations of tuberculous otitis media and common chronic suppurative otitis media. When met with the following conditions, we should pay highly attention to suspect tuberculous otitis media: The severity of local manifestations did not match with the length of the disease; with poor tympanic membrane at the early stage with no obvious cholesteatomas, with facial paralysis or hearing loss early onset; insensitive to traditional antibiotic treatment; with extensive granulation appeared in the tympanum and or mastoid cavity, with or without caseous necrosis or dead bone in the early days. The diagnosis should be confirmed based on the acid-fast staining of the histopathological section to detect positive acid-fast bacilli. Meanwhile, multiple laboratory examination methods（such as T-spot and PCR） should be integrated synchronously to help support the diagnosis.

Genomic signatures of climate adaptation in bank voles.

Evidence for divergent selection and adaptive variation across the landscape can provide insight into a species' ability to adapt to different environments. However, despite recent advances in genomics, it remains difficult to detect the footprints of climate-mediated selection in natural populations. Here, we analysed ddRAD sequencing data (21,892 SNPs) in conjunction with geographic climate variation to search for signatures of adaptive differentiation in twelve populations of the bank vole (Clethrionomys glareolus) distributed across Europe. To identify the loci subject to selection associated with climate variation, we applied multiple genotype-environment association methods, two univariate and one multivariate, and controlled for the effect of population structure. In total, we identified 213 candidate loci for adaptation, 74 of which were located within genes. In particular, we identified signatures of selection in candidate genes with functions related to lipid metabolism and the immune system. Using the results of redundancy analysis, we demonstrated that population history and climate have joint effects on the genetic variation in the pan-European metapopulation. Furthermore, by examining only candidate loci, we found that annual mean temperature is an important factor shaping adaptive genetic variation in the bank vole. By combining landscape genomic approaches, our study sheds light on genome-wide adaptive differentiation and the spatial distribution of variants underlying adaptive variation influenced by local climate in bank voles.

A pan-cancer multi-omics analysis of lactylation genes associated with tumor microenvironment and cancer development

BackgroundLactylation is a significant post-translational modification bridging the gap between cancer epigenetics and metabolic reprogramming. However, the association between lactylation and prognosis, tumor microenvironment (TME), and response to drug therapy in various cancers remains unclear. MethodsFirst, the expression, prognostic value, and genetic and epigenetic alterations of lactylation genes were systematically explored in a pan-cancer manner. Lactylation scores were derived for each tumor using the single-sample gene set enrichment analysis (ssGSEA) algorithm. The correlation of lactylation scores with clinical features, prognosis, and TME was assessed by integrating multiple computational methods. In addition, GSE135222 data was used to assess the efficacy of lactylation scores in predicting immunotherapy outcomes. The expression of lactylation genes in breast cancers and gliomas were verified by RNA-sequencing. ResultsLactylation genes were significantly upregulated in most cancer types. CREBBP and EP300 exhibited high mutation rates in pan-cancer analysis. The prognostic impact of the lactylation score varied by tumor type, and lactylation score was a protective factor for KIRC, ACC, READ, LGG, and UVM, and a risk factor for CHOL, DLBC, LAML, and OV. In addition, a high lactylation score was associated with cold TME. The infiltration levels of CD8+ T, γδT, natural killer T cell (NKT), and NK cells were lower in tumors with higher lactylation scores. Finally, immunotherapy efficacy was worse in patients with high lactylation scores than other types. ConclusionLactylation genes are involved in malignancy formation. Lactylation score serves as a promising biomarker for predicting patient prognosis and immunotherapy efficacy.

Analysis of the Effect Economic Growth, Open Unemployment Rate, Education, and PAD on Poverty Rate in Lumajang District

Poverty is a global problem faced by many countries in the world, including Indonesia. In an effort to achieve the Sustainable Development Goals (SDGs), poverty reduction is the main focus. This study aims to determine the effect of economic growth, unemployment rate, education, and local revenue on the poverty rate, and analyse the factors that influence poverty in Lumajang Regency. Multiple linear regression analysis method using Eviews 10 analysis tool was used in this study, with the research period covering 2012-2021. The results show that the Local Own Revenue (PAD) and Education Index have a significant influence on the poverty rate, while economic growth and unemployment rate have no significant influence on poverty. The findings suggest that the influence of unemployment, PAD, education, and economic growth on poverty rates in Indonesia is significant.

Software component selection methods and techniques: a systematic review

<p>Software component selection is critical in software engineering due to its vital role in reducing software development cost and time. This study analyzes software component selection research studies on methodologies, criteria, and multi-criteria decision-making (MCDM) techniques. The key study findings are: first, comprehensive standardized criteria for software component selection are lacking, with ambiguous terminology used in research. Second, current ad hoc selection processes need streamlining to reduce time, cost, and effort. Thus, an integrated approach is required to aid decision-makers. The review suggests developing automated tools or decision support systems combining multiple criteria decision methods to improve selection accuracy and efficiency. Standardized criteria catalogs can also assist software developers in the evaluation. The findings highlight that despite extensive academic research, component selection in practice remains sub-optimal. By informing future research and tool development, this review can benefit practitioners to systematically select the most appropriate software components meeting software requirements.</p>

Recent Advances in Conventional and Deep Learning-Based Depth Completion: A Survey.

Depth completion aims to recover pixelwise depth from incomplete and noisy depth measurements with or without the guidance of a reference RGB image. This task attracted considerable research interest due to its importance in various computer vision-based applications, such as scene understanding, autonomous driving, 3-D reconstruction, object detection, pose estimation, trajectory prediction, and so on. As the system input, an incomplete depth map is usually generated by projecting the 3-D points collected by ranging sensors, such as LiDAR in outdoor environments, or obtained directly from RGB-D cameras in indoor areas. However, even if a high-end LiDAR is employed, the obtained depth maps are still very sparse and noisy, especially in the regions near the object boundaries, which makes the depth completion task a challenging problem. To address this issue, a few years ago, conventional image processing-based techniques were employed to fill the holes and remove the noise from the relatively dense depth maps obtained by RGB-D cameras, while deep learning-based methods have recently become increasingly popular and inspiring results have been achieved, especially for the challenging situation of LiDAR-image-based depth completion. This article systematically reviews and summarizes the works related to the topic of depth completion in terms of input modalities, data fusion strategies, loss functions, and experimental settings, especially for the key techniques proposed in deep learning-based multiple input methods. On this basis, we conclude by presenting the current status of depth completion and discussing several prospects for its future research directions.

Predicting absolute risk for a person with missing risk factors.

We compared methods to project absolute risk, the probability of experiencing the outcome of interest in a given projection interval accommodating competing risks, for a person from the target population with missing predictors. Without missing data, a perfectly calibrated model gives unbiased absolute risk estimates in a new target population, even if the predictor distribution differs from the training data. However, if predictors are missing in target population members, a reference dataset with complete data is needed to impute them and to estimate absolute risk, conditional only on the observed predictors. If the predictor distributions of the reference data and the target population differ, this approach yields biased estimates. We compared the bias and mean squared error of absolute risk predictions for seven methods that assume predictors are missing at random (MAR). Some methods imputed individual missing predictors, others imputed linear predictor combinations (risk scores). Simulations were based on real breast cancer predictor distributions and outcome data. We also analyzed a real breast cancer dataset. The largest bias for all methods resulted from different predictor distributions of the reference and target populations. No method was unbiased in this situation. Surprisingly, violating the MAR assumption did not induce severe biases. Most multiple imputation methods performed similarly and were less biased (but more variable) than a method that used a single expected risk score. Our work shows the importance of selecting predictor reference datasets similar to the target population to reduce bias of absolute risk predictions with missing risk factors.

Education Research: Educating Child Neurology Residents About Psychogenic Nonepileptic Seizures

Background and Objectives Psychogenic nonepileptic seizures (PNES) are difficult to differentiate from epileptic seizures (ES) even for neurologists who see these conditions frequently. This difficulty is due to overlapping semiologic findings between the 2 diagnoses. Previous studies have shown that trainees, including neurology trainees, are not accurate in differentiating PNES from ES. Neurologists often find the communication of PNES difficult. Despite these challenges, most programs do not have formal curricula for teaching PNES, and there are no standards for residency curricula in this topic. The aim of this study was to understand the gaps in resident education on PNES. Methods This study was accomplished through a needs assessment of current pediatric neurology residents and residency program directors (PDs). Two unique surveys were distributed, 1 for child neurology trainees and 1 for PDs. Questions were designed to understand trainees' self-reported knowledge, confidence, current education received, and desired teaching. Similarly for PDs, questions were designed to assess the state of education on PNES at their programs, sufficiency of education, and their desire for standardized curriculum. Results Sixty-eight trainees and 21 PDs responded to the survey. Approximately one-quarter of trainees reported neutral to low levels of confidence and 38% reported neutral to low levels of knowledge in caring for patients with PNES. Trainees reported that directing patients with PNES to appropriate management was the most challenging aspect of care, followed by communicating the diagnosis, with 60% and 46% reporting difficulty, respectively. Only 21% of residents felt their current PNES education needs no improvement. One-fifth of PDs felt their current PNES education is not sufficient, and all reported they would incorporate a standardized curriculum. Trainees reported preferring to learn about this topic through lectures and simulation, while PDs preferred online modules and simulation. Discussion While residents and PDs report high confidence and knowledge in treating pediatric patients with PNES, respondents felt improvement is needed to their curricula regarding this topic. Multiple learning methods are preferred, with emphasis on communicating the diagnosis and management of patients once the diagnosis has been made. PDs desire a standardized curriculum and would incorporate one into their programs. Findings of this study could be used to create a national curriculum.

Attribution discernment of climate change and human interventions to runoff decline in Huangshui River Basin, China

Abstract To achieve sustainable development goals in Huangshui River Basin (HRB), strengthening adaptive water resources management under the dual impact of climate change (CC) and human interventions (HI) is of great significance. Multiple mathematical and statistical methods were employed to determine the runoff trend and breakpoint in HRB. The elasticity of CC and HI on the runoff decline and their contributions were quantitatively discerned based on the Budyko hypothesis, complementary method, and SWAT hydrological model. The results show that (1) the runoff showed a decreasing trend, with a runoff breakpoint in 1990; (2) the elasticity coefficients indicated a 1% increase in P, ET0, and n, leading to a 2.19% increase, a 1.19% decrease, and a 1.52% decrease in the runoff, respectively; (3) the Budyko framework determined the contribution of CC and HI to runoff decline in HRB to be 37.98–41.86% and 58.14–62.02%, respectively, and that estimated by SWAT hydrological model to be 38.72 and 61.28%, respectively; (4) HI were the primary factor for runoff decline in HRB, where direct anthropogenic disturbances such as water withdrawals and water conservancy project construction were the main drivers. The findings have important scientific significance for water resources planning and management in HRB.

Profiling of Oxylipins as Markers of Oxidative Stress in Biological Samples.

Oxylipins are oxidized metabolites of polyunsaturated fatty acids (PUFAs). They represent a class of risk markers and/or therapeutic targets for diseases associated with inflammation, including cardiovascular disease and brain disorders. Because the biological activities of free PUFAs and oxylipins depend on their chemical structures and concentrations, monitoring PUFAs and oxylipin levels in biological systems is critical for understanding their roles in health and disease. Traditionally, accurate quantification of free PUFAs and oxylipins in biological samples was performed separately, as PUFAs are often 1000-fold more abundant than the derived oxidized fatty acids (oxylipins). This article describes a liquid chromatography multiple reaction monitoring tandem mass spectrometry method for the quantitative analysis of five free PUFAs and 88 oxylipins in various biological fluids, including plasma, platelet supernatants, and tissues. The same approach can also be used in conjunction with an alkaline hydrolysis step to quantify total oxylipins in fish oils. We observed that in some samples, linoleic acid levels in plasma and eicosapentaenoic acid and arachidonic acid levels in brain tissue were above the upper limit of quantification. To address this issue, we developed a data analysis method to obtain PUFA and oxylipin concentrations in these samples without additional sample preparation, thus significantly saving time and labor. © 2024 Wiley Periodicals LLC. Basic Protocol: Quantification of polyunsaturated fatty acids (PUFAs) and oxylipins using liquid chromatography multiple reaction monitoring tandem mass spectrometry Support Protocol 1: Preparation of internal standard mixed working solution Support Protocol 2: Preparation of standard mixed stock solution Support Protocol 3: Preparation of standard mixed working solution Alternate Protocol 1: Extraction and quantitation of free PUFAs and oxylipins from mouse brain tissue Alternate Protocol 2: Extraction and quantitation of total PUFAs and oxylipins from fish oil.

One level deep convolutional neural network for facial key points detection

Facial landmark detection has a lot of applications in face recognition, face alignment, facial expression recognition, video surveillance and security systems. In the existing literature, there are multiple methods utilizing convolutional neural networks (CNNs) that address this problem in various ways. In many cases, the models use a tree-like structure of CNNs to achieve better results. This paper proposes a combination of three parallel deep convolutional neural networks (DCNNs) to estimate the accurate localization of each keypoint. The first one focuses on the whole face to outperform five points, including the eyes, nose, and mouth corners. The second one focuses on the eyes-nose parts to outperform three points, specifically the eyes and nose. The last one focuses on the nose-mouth parts to outperform three points, namely the nose and mouth corners. Further, we combine all outputs of the three DCNNs and take the average value of each detected key point as the final output. In the first step, we improvthe the parameter efficiency and accuracy of each DCNNs through a set of experiments using the labeled face parts in-the-wild database (LFPW) and the helen facial feature dataset (Helen). Then, we demonstrate that our approach yields more accurate estimations of facial key points than two state-of-the-art methods and commercial software in terms of accuracy.

A Comprehensive Review on Machine Learning Techniques for Protein Family Prediction.

Proteomics is a field dedicated to the analysis of proteins in cells, tissues, and organisms, aiming to gain insights into their structures, functions, and interactions. A crucial aspect within proteomics is protein family prediction, which involves identifying evolutionary relationships between proteins by examining similarities in their sequences or structures. This approach holds great potential for applications such as drug discovery and functional annotation of genomes. However, current methods for protein family prediction have certain limitations, including limited accuracy, high false positive rates, and challenges in handling large datasets. Some methods also rely on homologous sequences or protein structures, which introduce biases and restrict their applicability to specific protein families or structures. To overcome these limitations, researchers have turned to machine learning (ML) approaches that can identify connections between protein features and simplify complex high-dimensional datasets. This paper presents a comprehensive survey of articles that employ various ML techniques for predicting protein families. The primary objective is to explore and improve ML techniques specifically for protein family prediction, thus advancing future research in the field. Through qualitative and quantitative analyses of ML techniques, it is evident that multiple methods utilizing a range of classifiers have been applied for protein family prediction. However, there has been limited focus on developing novel classifiers for protein family classification, highlighting the urgent need for improved approaches in this area. By addressing these challenges, this research aims to enhance the accuracy and effectiveness of protein family prediction, ultimately facilitating advancements in proteomics and its diverse applications.

Developing Student Expertise in Evolution: Cognitive Construals Complement Key Concepts in Student Representations.

Genetic variation is historically challenging for undergraduate students to master, potentially due to its grounding in both evolution and genetics. Traditionally, student expertise in genetic variation has been evaluated using Key Concepts. However, Cognitive Construals may add to a more nuanced picture of students' developing expertise. Here, we analyze the occurrence of Key Concepts and Cognitive Construals among three types of student representations: interviews, drawn models, and constructed responses (CRs). Our mixed-methods analysis indicates that differential survival and differential reproduction occur more often in interviews than in CRs. In our interviews, presence of Cognitive Construals indicate varying levels of understanding of genetic variation, but we were not able to detect Cognitive Construals in students' models or CRs. Finally, our analyses of both Key Concepts and Cognitive Construals in student representations indicate that Cognitive Construals can co-occur with any number of Key Concepts, and that the presence of Construal-based language alone does not seem to correlate to the expert nature of a response. Taken together, our results highlight the need for instructors to avoid treating Construal-based language as implicit disconnects in student understanding, and to use multiple methods to gain a holistic picture of student expertise.

Risk factors for violence in an emergency department: Nurses' perspectives.

Work-related violence remains a significant problem in healthcare settings, including EDs. Violence risk assessment tools have been developed to improve risk mitigation in this setting; however, incorporation of these tools into standard hospital processes remains scarce. This research aimed to explore nurses' perspectives on the Bröset Violence Checklist used in routine violence risk assessment and their recommendations for additional items. Thirty nursing staff who used the Bröset Violence Checklist (BVC) as standard practice for 5 years participated in two focus groups where 23 violence risk factors were presented. Using multiple methods, participants were asked to select and elaborate from a pre-determined list what they considered most useful in violence risk assessment in respect to descriptors and terminology. Quantitative data showed most risk factors presented to the group were considered to be predictive of violence. Ten were regarded as associated with risk, and overt behaviours received the highest votes. The terms 'shouting and demanding' was preferred over 'boisterous', and 'cognitive impairment' over 'confusion'. Patient clinical characteristics and staff perceptions of harm, inability to observe subtle behaviour, imposed restrictions and interventions and environmental conditions and impact were also important considerations. We recommend that violence risk assessment include: history of violence, cognitive impairment, psychotic symptoms, drug and alcohol influence, shouting and demanding, verbal abuse/hostility, impulsivity, agitation, irritability and imposed restrictions and interventions. These violence risk factors fit within the four categories of historical, clinical, behavioural and situational.

Seasonal and interannual dissolved organic carbon transport process dynamics in a subarctic headwater catchment revealed by high-resolution measurements

Abstract. Dissolved organic carbon (DOC) dynamics are evolving in the rapidly changing Arctic and a comprehensive understanding of the controlling processes is urgently required. For example, the transport processes governing DOC dynamics are prone to climate-driven alteration given their strong seasonal nature. Hence, high-resolution and long-term studies are required to assess potential seasonal and interannual changes in DOC transport processes. In this study, we monitored DOC at a 30 min resolution from September 2018 to December 2022 in a headwater peatland-influenced stream in northern Finland (Pallas catchment, 68° N). Temporal variability in transport processes was assessed using multiple methods: concentration–discharge (C–Q) slope for seasonal analysis, a modified hysteresis index for event analysis, yield analysis, and random forest regression models to determine the hydroclimatic controls on transport. The findings revealed the following distinct patterns: (a) the slope of the C–Q relationship displayed a strong seasonal trend, indicating increasing transport limitation each month after snowmelt began; (b) the hysteresis index decreased post-snowmelt, signifying the influence of distal sources and DOC mobilization through slower pathways; and (c) interannual variations in these metrics were generally low, often smaller than month-to-month fluctuations. These results highlight the importance of long-term and detailed monitoring to enable separation of inter- and intra-annual variability to better understand the complexities of DOC transport. This study contributes to a broader comprehension of DOC transport dynamics in the Arctic, specifically quantifying seasonal variability and associated mechanistic drivers, which is vital for predicting how the carbon cycle is likely to change in Arctic ecosystems.

A Perspective of Theory Technology-To-Performance Chain (TPC): Accounting Information Systems, Internal Control and Work Motivation on Employee Performance

The advancement of technology is a reflection of the contemporary era, continually evolving to address the changing needs of humanity. Beyond the widespread demand for accounting information systems, it is crucial to recognize the substantial influence that internal control and work motivation wield over employee performance. This research endeavors to examine the impact of accounting information systems, internal control, and work motivation on the performance of PT Arfia Megah's employees. The study employs a saturated sampling method, encompassing the entire population of 40 employees at PT Arfia Megah. The independent variables under scrutiny include accounting information systems, internal controls, and work motivation, while employee performance serves as the dependent variable. Utilizing a quantitative approach, primary data is collected through a questionnaire, and the analysis is conducted using the multiple regression method. The findings affirm a positive correlation between accounting information systems, internal control, work motivation, and enhanced employee performance

Nonlinear Dynamic Analysis of FG Fluid Conveying Micropipes with Initial Imperfections

This paper investigates the nonlinear dynamics of FG-FCMPs with initial imperfections. Based on the MCST, the nonlinear equations of motion and the corresponding boundary conditions are established by applying Hamilton’s principle, the Euler–Bernoulli beam theory, and von-Kármán geometric nonlinearity. To describe the initial geometric imperfection of the FG-FCMP, the first-order vibrational mode is employed. Subsequently, in cases of primary parametric resonance, 1:2 subharmonic resonance for the first-order mode, as well as primary resonance for the second-order mode, Galerkin’s method, and the multiple scale method are utilized to analyze the amplitude–frequency responses of the imperfect FG-FCMP. The numerical simulations test the influence of flow velocity, micro-scale parameter, geometrical imperfections, and external loads on the nonlinear characteristics of a coupled system with two DOFs. It is found that, as the increase of flow velocity, micro-scale effect, and external load, the amplitude of the first two modes can be increased. The hardening characteristics are converted into the softening characteristics due to the imperfect effect. Furthermore, numerical results provide a more comprehensive understanding of the nonlinear dynamics of FCMPs for both periodic and chaotic motions.

The Effect of Good Corporate Governance (GCG) on Financial Performance with Corporate Social Responsibility (CSR) as an Intervening Variable (Empirical Study on Primary Consumer Sub Sector Companies Listed on The IDX 2018 - 2022)

This research aims to examine the influence of Good Corporate Governance (GCG) including institutional ownership and the board of commissioners on financial performance with Corporate Social Responsibility (CSR) as an intervening variable. This research uses financial performance as the dependent variable and Good Corporate Governance as a proxy for institutional ownership and the board of commissioners as the independent variable and Corporate Social Responsibility (CSR) as the intervening variable. This type of research is associative research. This research uses secondary data in the form of annual financial reports originating from the Indonesia Stock Exchange (BEI). The population in this research is Primary Consumer Sub-Sector Companies listed on the Indonesia Stock Exchange 2018-2022. The sampling technique in this research used a purposive sampling method, 7 companies were obtained as research samples. The data analysis technique in this study uses the multiple linear regression analysis method and the data analysis tool in this study uses the Eviews version 12 software program. The results of the study show that GCG and CSR together have a positive effect on the company's financial performance, while the proxies for institutional ownership and the board commissioner is partially insignificant. Likewise, CSR does not have a significant impact on financial performance. In addition, the board of commissioners proxy influences financial performance through CSR as a moderation, while the institutional ownership proxy does not have a similar effect.Keywords: Good Corporate Governance, Institutional Ownership, Board of Commissioners, Corporate Social Responsibility, Financial Performance

Automatic Prediction of Peak Optical Absorption Wavelengths in Molecules Using Convolutional Neural Networks.

Molecular design depends heavily on optical properties for applications such as solar cells and polymer-based batteries. Accurate prediction of these properties is essential, and multiple predictive methods exist, from ab initio to data-driven techniques. Although theoretical methods, such as time-dependent density functional theory (TD-DFT) calculations, have well-established physical relevance and are among the most popular methods in computational physics and chemistry, they exhibit errors that are inherent in their approximate nature. These high-throughput electronic structure calculations also incur a substantial computational cost. With the emergence of big-data initiatives, cost-effective, data-driven methods have gained traction, although their usability is highly contingent on the degree of data quality and sparsity. In this study, we present a workflow that employs deep residual convolutional neural networks (DR-CNN) and gradient boosting feature selection to predict peak optical absorption wavelengths (λmax) exclusively from SMILES representations of dye molecules and solvents; one would normally measure λmax using UV-vis absorption spectroscopy. We use a multifidelity modeling approach, integrating 34,893 DFT calculations and 26,395 experimentally derived λmax data, to deliver more accurate predictions via a Bayesian-optimized gradient boosting machine. Our approach is benchmarked against the state of the art that is reported in the scientific literature; results demonstrate that learnt representations via a DR-CNN workflow that is integrated with other machine learning methods can accelerate the design of molecules for specific optical characteristics.