Unstructured Variability Research Articles

The Application of Fuzzy Time Series Method for Web-Based Prediction of Household Chemical Product Sales Stock

The seasonal and trend factors of the market demand greatly influence the sales of household chemical products. Therefore, PT. XYZ company requires an accurate sales prediction system to optimize production and inventory. This research aims to develop a web-based sales prediction system for household chemical products using fuzzy time series method. This method considers the factors that influence sales and generates a prediction model that can assist the company in decision-making. This method allows the use of uncertain and unstructured variables in the prediction. The research collected historical sales data of household chemical products from January to December 2023 through interviews, observations, and literature studies. The system was designed using a structured approach and PHP programming language. The Fuzzy Time Series method was used for decision-making. The accuracy of the Fuzzy Time Series method for sales data from January to December 2020 was calculated through program simulation and manual methods, resulting in an accuracy rate of 13.71%. The application provides information on the sales of each product, with the hope that users can reduce the accumulation of stock every month.

Using Dynamic Bayesian Optimization to Induce Desired Effects in the Presence of Motor Learning: a Simulation Study.

Human-in-the-loop (HIL) optimization is a control paradigm used for tuning the control parameters of human-interacting devices while accounting for variability among individuals. A limitation of state-of-the-art HIL optimization algorithms such as Bayesian Optimization (BO) is that they assume that the relationship between control parameters and user response does not change over time. BO can be modified to account for the dynamics of the user response by implementing time into the kernel function, a method known as Dynamic Bayesian Optimization (DBO). However, it is unknown if DBO outperforms BO when the human response is characterized by models of human motor learning. In this work, we simulated runs of HIL optimization using BO and DBO towards establishing if DBO is a suitable paradigm for HIL optimization in the presence of motor learning. Simulations were conducted assuming either purely time-dependent participant responses, or assuming that responses would arise from state-space models of motor learning capable of describing both adaptation and use-dependent learning behavior. Statistical comparisons indicated that DBO was never inferior to BO, and, after a certain number of iterations, generally outperformed BO in convergence to optimal inputs and outputs. The number of iterations beyond which DBO was superior to BO occurred earlier when the input-output relationship of the simulated responses was more dynamic. Our results suggest that DBO may improve the performance of HIL optimization over BO when a sufficient number of iterations can be evaluated to accurately distinguish between unstructured variability (noise) and learning.

Machine Learning model based on the Light Gradient Boosting Machine to predict the probability of default in customers of the Credit Card portfolio

Bank loans are a widely used means of payment in recent times, more and more people are accessing products such as credit cards, loans, etc. Banks have implemented classic prediction models, the vast majority based on logistic regression since it allows great interpretability for the business and the effect of the model variables. The purpose of this research is to perform a predictive analysis on the probability of customer default in the credit card portfolio using a risk score. The dataset used is the so-called default of credit card clients Data Set from the UCI Machine Learning DB, the approach is quantitative and the methodology is descriptive analytics, techniques based on gradient boosters will be used to make the prediction, among the trained algorithms We have Logistic Regression with WOE, CatBoost, As a result, the light gradient enhancement machine (LightGBM) tuned with a Bayesian search was obtained, obtaining a GINI of 57.4, which improves by +6 points to the Logistic Regression with Woe and by +3p to XgBoost and CatBoost. Finally, obtaining the Gain and Shapley values made up for the lack of interpretability of the variables, allowing better decision making when evaluating clients. Likewise, as future work, it is intended to add unstructured variables that allow the Model's indicators to be improved.

Prediction of acute appendicitis among patients with undifferentiated abdominal pain at emergency department

BackgroundEarly screening and accurately identifying Acute Appendicitis (AA) among patients with undifferentiated symptoms associated with appendicitis during their emergency visit will improve patient safety and health care quality. The aim of the study was to compare models that predict AA among patients with undifferentiated symptoms at emergency visits using both structured data and free-text data from a national survey.MethodsWe performed a secondary data analysis on the 2005-2017 United States National Hospital Ambulatory Medical Care Survey (NHAMCS) data to estimate the association between emergency department (ED) patients with the diagnosis of AA, and the demographic and clinical factors present at ED visits during a patient’s ED stay. We used binary logistic regression (LR) and random forest (RF) models incorporating natural language processing (NLP) to predict AA diagnosis among patients with undifferentiated symptoms.ResultsAmong the 40,441 ED patients with assigned International Classification of Diseases (ICD) codes of AA and appendicitis-related symptoms between 2005 and 2017, 655 adults (2.3%) and 256 children (2.2%) had AA. For the LR model identifying AA diagnosis among adult ED patients, the c-statistic was 0.72 (95% CI: 0.69–0.75) for structured variables only, 0.72 (95% CI: 0.69–0.75) for unstructured variables only, and 0.78 (95% CI: 0.76–0.80) when including both structured and unstructured variables. For the LR model identifying AA diagnosis among pediatric ED patients, the c-statistic was 0.84 (95% CI: 0.79–0.89) for including structured variables only, 0.78 (95% CI: 0.72–0.84) for unstructured variables, and 0.87 (95% CI: 0.83–0.91) when including both structured and unstructured variables. The RF method showed similar c-statistic to the corresponding LR model.ConclusionsWe developed predictive models that can predict the AA diagnosis for adult and pediatric ED patients, and the predictive accuracy was improved with the inclusion of NLP elements and approaches.

Multi-resolution B-splines data compression improves MIR spectroscopy-based Health diagnostic efficiency

MIR spectroscopy is becoming an increasingly important tool potentially useful for diagnosis purposes especially by studying body fluids. Indeed, diseases induce changes in the composition of fluids modifying the MIR spectra. However, such changes can be difficult to capture if the structure of the data is not considered. Our objective was to improve MIR spectra analysis by using approximation of the spectra by B-splines at different specific resolutions and to combine these spectra representations with a machine learning model to predict hepatic steatosis from serum study. The different resolutions make it possible to identify changes in shape over bands of various widths. The multiresolution model helps to improve the hepatic steatosis prediction compared to conventional approaches where the absorbances are considered as unstructured variables. In addition, B-splines provide both localized and compressed information that can be translated into biochemical terms more easily than with other classical approximation methods (wavelets, Fourier transforms).

Towards activity-based exposure measures in spatial analysis of pedestrian–motor vehicle crashes

BackgroundAlthough numerous efforts have been devoted to exploring the effects of area-wide factors on the frequency of pedestrian crashes in neighborhoods over the past two decades, existing studies have largely failed to provide a full picture of the factors that contribute to the incidence of zonal pedestrian crashes, due to the unavailability of reliable exposure data and use of less sound analytical methods. MethodsBased on a crowdsourced dataset in Hong Kong, we first proposed a procedure to extract pedestrian trajectories from travel-diary survey data. We then aggregated these data to 209 neighborhoods and developed a Bayesian spatially varying coefficients model to investigate the spatially non-stationary relationships between the number of pedestrian–motor vehicle (PMV) crashes and related risk factors. To dissect the role of pedestrian exposure, the estimated coefficients of models with population, walking trips, walking time, and walking distance as the measure of pedestrian exposure were presented and compared. ResultsOur results indicated substantial inconsistencies in the effects of several risk factors between the models of population and activity-based exposure measures. The model using walking trips as the measure of pedestrian exposure had the best goodness-of-fit. We also provided new insights that in addition to the unstructured variability, heterogeneity in the effects of explanatory variables on the frequency of PMV crashes could also arise from the spatially correlated effects. After adjusting for vehicle volume and pedestrian activity, road density, intersection density, bus stop density, and the number of parking lots were found to be positively associated with PMV crash frequency, whereas the percentage of motorways and median monthly income had negative associations with the risk of PMV crashes. ConclusionsThe use of population or population density as a surrogate for pedestrian exposure when modeling the frequency of zonal pedestrian crashes is expected to produce biased estimations and invalid inferences. Spatial heterogeneity should also not be negligible when modeling pedestrian crashes involving contiguous spatial units.

Use of natural language processing to improve predictive models for imaging utilization in children presenting to the emergency department

ObjectiveTo examine the association between the medical imaging utilization and information related to patients’ socioeconomic, demographic and clinical factors during the patients’ ED visits; and to develop predictive models using these associated factors including natural language elements to predict the medical imaging utilization at pediatric ED.MethodsPediatric patients’ data from the 2012–2016 United States National Hospital Ambulatory Medical Care Survey was included to build the models to predict the use of imaging in children presenting to the ED. Multivariable logistic regression models were built with structured variables such as temperature, heart rate, age, and unstructured variables such as reason for visit, free text nursing notes and combined data available at triage. NLP techniques were used to extract information from the unstructured data.ResultsOf the 27,665 pediatric ED visits included in the study, 8394 (30.3%) received medical imaging in the ED, including 6922 (25.0%) who had an X-ray and 1367 (4.9%) who had a computed tomography (CT) scan. In the predictive model including only structured variables, the c-statistic was 0.71 (95% CI: 0.70–0.71) for any imaging use, 0.69 (95% CI: 0.68–0.70) for X-ray, and 0.77 (95% CI: 0.76–0.78) for CT. Models including only unstructured information had c-statistics of 0.81 (95% CI: 0.81–0.82) for any imaging use, 0.82 (95% CI: 0.82–0.83) for X-ray, and 0.85 (95% CI: 0.83–0.86) for CT scans. When both structured variables and free text variables were included, the c-statistics reached 0.82 (95% CI: 0.82–0.83) for any imaging use, 0.83 (95% CI: 0.83–0.84) for X-ray, and 0.87 (95% CI: 0.86–0.88) for CT.ConclusionsBoth CT and X-rays are commonly used in the pediatric ED with one third of the visits receiving at least one. Patients’ socioeconomic, demographic and clinical factors presented at ED triage period were associated with the medical imaging utilization. Predictive models combining structured and unstructured variables available at triage performed better than models using structured or unstructured variables alone, suggesting the potential for use of NLP in determining resource utilization.

Development of an Algorithm to Identify Patients with Physician-Documented Insomnia

We developed an insomnia classification algorithm by interrogating an electronic medical records (EMR) database of 314,292 patients. The patients received care at Massachusetts General Hospital (MGH), Brigham and Women’s Hospital (BWH), or both, between 1992 and 2010. Our algorithm combined structured variables (such as International Classification of Diseases 9th Revision [ICD-9] codes, prescriptions, laboratory observations) and unstructured variables (such as text mentions of sleep and psychiatric disorders in clinical narrative notes). The highest classification performance of our algorithm was achieved when it included a combination of structured variables (billing codes for insomnia, common psychiatric conditions, and joint disorders) and unstructured variables (sleep disorders and psychiatric disorders). Our algorithm had superior performance in identifying insomnia patients compared to billing codes alone (area under the receiver operating characteristic curve [AUROC] = 0.83 vs. 0.55 with 95% confidence intervals [CI] of 0.76–0.90 and 0.51–0.58, respectively). When applied to the 314,292-patient population, our algorithm classified 36,810 of the patients with insomnia, of which less than 17% had a billing code for insomnia. In conclusion, an insomnia classification algorithm that incorporates clinical notes is superior to one based solely on billing codes. Compared to traditional methods, our study demonstrates that a classification algorithm that incorporates physician notes can more accurately, comprehensively, and quickly identify large cohorts of insomnia patients.

Predicting Innovation Success in the Motion Picture Industry: The Influence of Multiple Quality Signals

In settings characterized by frequent product entries and rapid market exit, predicting innovation success is a difficult task. The challenge in such short life‐cycle markets is exacerbated when the offering is an experience good (i.e., product quality is not known prior to consumption). This article investigates a variety of information disequilibrium‐reducing cues that can signal a new product's quality, and posit that those cues affect knowledge of and attitude toward (i.e., customers' mindset) the item, and in turn, demand. The work draws on signaling theory in an experience good context (motion pictures) to set up a nomological net that includes three sources of quality cues as predictors of consumer demand, namely: (i) traditional prelaunch decisions (e.g., production budget, advertising budget, and sequel), (ii) volume and valence of movie critics' reviews, and (iii) text contained in the movie critics' reviews. Based on a stratified, random sample of movie introductions (n = 115), the results show that the prelaunch (structured numeric) variables alone explain a meaningful proportion of the variance in domestic box office sales. The postlaunch (structured numeric) variables, volume and valence of critics' reviews, add more explanatory power, and analyzing over two million words from the (unstructured) text of critics' reviews adds further explanatory power. This research answers a recent call to model the variety of structured and unstructured variables to predict innovation success in data‐rich environments, and thereby represents the pioneering study to combine all three sources of quality cues—(i) prelaunch, structured variables; (ii) postlaunch, structured variables; and (iii) postlaunch, unstructured variables derived from text reviews—as determinants of demand. The approach taken also substantiates film as innovation and as an interesting case of new product development. For managers in a short life‐cycle, experience good settings, the results support the inclusion of a broader “variety” of quality signals to improve forecasting in today's data‐rich environments.Practitioner PointsThe results here highlight the importance of considering both marketing decisions and expert reviews in forecasting approaches for new product launches. Exploring a nuanced approach to analyzing expert reviews seems to be particularly promising in predicting product successes and failures.This research also helps us develop a deeper understanding of the nature of critics' reviews. Beyond just “thumbs up/thumbs down” sort of recommendations, we identify specific words that are influential in making an impression on consumers and, ultimately, driving product success or failure.This work also emphasizes the challenges of predicting success in markets for short life‐cycle experience goods such as films. In these markets, expert reviews likely carry even more weight since consumers have great difficulty in assessing the product in advance.

Scaling impacts on environmental controls and spatial heterogeneity of soil organic carbon stocks

Abstract. The spatial heterogeneity of land surfaces affects energy, moisture, and greenhouse gas exchanges with the atmosphere. However, representing the heterogeneity of terrestrial hydrological and biogeochemical processes in Earth system models (ESMs) remains a critical scientific challenge. We report the impact of spatial scaling on environmental controls, spatial structure, and statistical properties of soil organic carbon (SOC) stocks across the US state of Alaska. We used soil profile observations and environmental factors such as topography, climate, land cover types, and surficial geology to predict the SOC stocks at a 50 m spatial scale. These spatially heterogeneous estimates provide a data set with reasonable fidelity to the observations at a sufficiently high resolution to examine the environmental controls on the spatial structure of SOC stocks. We upscaled both the predicted SOC stocks and environmental variables from finer to coarser spatial scales (s = 100, 200, and 500 m and 1, 2, 5, and 10 km) and generated various statistical properties of SOC stock estimates. We found different environmental factors to be statistically significant predictors at different spatial scales. Only elevation, temperature, potential evapotranspiration, and scrub land cover types were significant predictors at all scales. The strengths of control (the median value of geographically weighted regression coefficients) of these four environmental variables on SOC stocks decreased with increasing scale and were accurately represented using mathematical functions (R2 = 0.83–0.97). The spatial structure of SOC stocks across Alaska changed with spatial scale. Although the variance (sill) and unstructured variability (nugget) of the calculated variograms of SOC stocks decreased exponentially with scale, the correlation length (range) remained relatively constant across scale. The variance of predicted SOC stocks decreased with spatial scale over the range of 50 m to ~ 500 m, and remained constant beyond this scale. The fitted exponential function accounted for 98 % of variability in the variance of SOC stocks. We found moderately accurate linear relationships between mean and higher-order moments of predicted SOC stocks (R2 ∼ 0.55–0.63). Current ESMs operate at coarse spatial scales (50–100 km), and are therefore unable to represent environmental controllers and spatial heterogeneity of high-latitude SOC stocks consistent with observations. We conclude that improved understanding of the scaling behavior of environmental controls and statistical properties of SOC stocks could improve ESM land model benchmarking and perhaps allow representation of spatial heterogeneity of biogeochemistry at scales finer than those currently resolved by ESMs.

Incidence variation of prostate and cervical cancer according to socioeconomic level in the Girona Health Region.

BackgroundThe main aim of this study, using a spatial-temporal model, is to analyse the link between a deprivation index and the incidence of prostate and cervical cancer in the Girona Health Region (GHR).MethodsThis is a population-based study which includes all the inhabitants in the GHR in the period 1993–2006. In order to assess prostate/cervical cancer risk, Besag, York and Mollie (BYM)’s spatial-temporal version of the model was used and four random effects were introduced: (non-spatial) unstructured variability, spatial dependency, temporal dependency and spatial-temporal interaction. As an explanatory variable, a deprivation index was introduced at the census tract level. Furthermore, the percentage of the population between 45–64 years of age and over-65 was also considered as explanatory variables.ResultsIn the case of prostate cancer, all the variables which were introduced into the model showed a significant correlation with the relative risk, except for the second quintile of the deprivation index. Furthermore, as the index increased the correlation became negative and lower. Thus, the correlation between the relative risk and the two age bands proved to be lower, the higher the age was. In the case of cervical cancer, only the correlation between the over-65 age band and the relative risk was found to be statistically significant and positive.ConclusionsIn the case of prostate cancer, the results obtained in the GHR are in line with similar analyses. However, in the case of cervical cancer, no significant relationship between incidences in this location or economic status was found.

Improving the power of genetic association tests with imperfect phenotype derived from electronic medical records.

To reduce costs and improve clinical relevance of genetic studies, there has been increasing interest in performing such studies in hospital-based cohorts by linking phenotypes extracted from electronic medical records (EMRs) to genotypes assessed in routinely collected medical samples. A fundamental difficulty in implementing such studies is extracting accurate information about disease outcomes and important clinical covariates from large numbers of EMRs. Recently, numerous algorithms have been developed to infer phenotypes by combining information from multiple structured and unstructured variables extracted from EMRs. Although these algorithms are quite accurate, they typically do not provide perfect classification due to the difficulty in inferring meaning from the text. Some algorithms can produce for each patient a probability that the patient is a disease case. This probability can be thresholded to define case-control status, and this estimated case-control status has been used to replicate known genetic associations in EMR-based studies. However, using the estimated disease status in place of true disease status results in outcome misclassification, which can diminish test power and bias odds ratio estimates. We propose to instead directly model the algorithm-derived probability of being a case. We demonstrate how our approach improves test power and effect estimation in simulation studies, and we describe its performance in a study of rheumatoid arthritis. Our work provides an easily implemented solution to a major practical challenge that arises in the use of EMR data, which can facilitate the use of EMR infrastructure for more powerful, cost-effective, and diverse genetic studies.

Genetic diversity in Hemileia vastatrix based on RAPD markers

Random amplified polymorphic DNA (RAPD) was used to assess the genetic structure of Hemileia vastatrix populations. Forty-five rust isolates with different virulence spectra and from different hosts and geographical regions were analyzed. Out of 45 bands, generated with three RAPD primers, 35 (78%) were polymorphic and scored as molecular markers. Cluster analysis exhibits unstructured variability of this pathogen with regard to physiological race, geographical origin or host. The genotypic diversity (H′) inferred from Shannon’s index was higher than gene diversity (Ht), suggesting that diversity is distributed among clonal lineages. Estimates of gene diversity in Africa and Asia populations were higher in total (Ht) as compared to within population diversity (Hs). Genetic differentiation was considerable among coffee rust isolates from Africa (Gst = 0.865) and Asia (Gst = 0.768) but not among isolates from South America (Gst = 0.266). We concluded that genetic diversity in H. vastatrix was moderately low and that the genetic differentiation among populations shows that asexual reproduction is likely to play an important role in the population biology of this fungus. This should be taken into account for the development of breeding programs.

Genetic diversity in Hemileia vastatrix based on RAPD markers

Random amplified polymorphic DNA (RAPD) was used to assess the genetic structure of Hemileia vastatrix populations. Forty-five rust isolates with different virulence spectra and from different hosts and geographical regions were analyzed. Out of 45 bands, generated with three RAPD primers, 35 (78%) were polymorphic and scored as molecular markers. Cluster analysis exhibits unstructured variability of this pathogen with regard to physiological race, geographical origin or host. The genotypic diversity (H') inferred from Shannon's index was higher than gene diversity (Ht), suggesting that diversity is distributed among clonal lineages. Estimates of gene diversity in Africa and Asia populations were higher in total (Ht) as compared to within population diversity (Hs). Genetic differentiation was considerable among coffee rust isolates from Africa (Gst = 0.865) and Asia (Gst = 0.768) but not among isolates from South America (Gst = 0.266). We concluded that genetic diversity in H. vastatrix was moderately low and that the genetic differentiation among populations shows that asexual reproduction is likely to play an important role in the population biology of this fungus. This should be taken into account for the development of breeding programs.

Inter-Simple Sequence Repeat fingerprints to assess genetic diversity in Tunisian fig (Ficuscarica L.) germplasm

The genetic diversity of 18 Tunisian fig cultivars was investigated at the DNA level using the Inter Simple Sequence Repeat (ISSR) associated with the Polymerase Chain Reaction (PCR). Using a set of primers, the most informative ones were selected that were characterized by an important Resolving power value of 29.6. A total of 47 discernible fragments were scored from samples, with a mean of 11.7 fragments per primer. The 90.4% of sample that were polymorphic were scored as molecular markers to examine the Tunisian fig germplasm polymorphism at DNA level. A large genetic diversity as related to ISSR patterns was found within the local Tunisian fig germplasm. An UPGMA dendrogram exhibits the unstructured variability in this crop. Moreover, the principal component analysis shows that the observed diversity was typically continuous. Our data provide a large number of ISSR markers that are useful in the fingerprinting of Ficus carica L. cultivars, and in the understanding of the genetic relationships among these accessions.

Bayesian models for spatial incidence: a case study of suicide using the BUGS program

The present study considers variability in suicidal outcomes (attempted and completed suicide) over small areas in London. These small areas, electoral wards, average just under 10 thousand in population. A Bayesian framework allowing for prior uncertainty in structured and unstructured variability (respectively spatially correlated errors and pure heterogeneity) is adopted. Covariates representing the summary effects of deprivation and anomie (social disorganization) are included in the analysis, and their contextual variability modelled in the same way as variability in the level of suicidal outcomes.

The future of sensors for intelligent machine control

The progress made towards solving the problem of applying robotics for small and medium batch production since 1960s has been very limited. By analysing the strengths and weaknesses of robot technology some practical rules for robotic application have been previously developed by the author. These 'laws' are further considered within the context of the needs of 'sensory' systems as an aid to robotic machinery in order to throw some light on the following: (a) the future practical architecture of robotic an sensor systems and (b) the direction for research and development to take. Most problems are concerned with the better structuring of material and component flow to and from a robotic workstation and the reduction of unstructured variability at the workplace. This is to enable sensors to be applied in a viable commercial manner.

Screen-oriented highlevel debugger(SHD) for PASCAL

There are certain golden maxims in programming. One of them is, that no matter how meticulous a programmer is, he spends major portion of his software development time on debugging. In such wise, we must have good debuggers to reduce debugging time. As Lauesen[1] puts it, the superiority of a debugger lies in helping user catch more bugs in given time. Accordingly, power of a debugger is the sole yardstick for goodness of it. Power in this context mean both, functionality and convenience. Other factors, such as, extra information it requires from compiler, its bigger size etc. are irrelevant so long as it is extra powerful. The compexity of the debugger, in terms of efforts required to program it should not be a prohibiting factor either. A debugger is used, as often as, if not more than the compiler itself. Conventional Debuggers[2] which call themselves 'High Level Debugger' are no doubt better than 'Dumpers' but are still too primitive to help user in any substantial manner. Most of them are confined to setting up of breakpoints and examination of unstructured variables. SHD strives to match its level of commands with the thought process of the user for maximum efficiency in detecting bugs.

Screen-oriented highlevel debugger (SHD) for Pascal

There are certain golden maxims in programming. One of them is, that no matter how meticulous a programmer is, he spends major portion of his software development time on debugging. In such wise, we must have good debuggers to reduce debugging time. As Lauesen[1] puts it, the superiority of a debugger lies in helping user catch more bugs in given time. Accordingly, power of a debugger is the sole yardstick for goodness of it. Power in this context mean both, functionality and convenience. Other factors, such as, extra information it requires from compiler, its bigger size etc. are irrelevant so long as it is extra powerful. The compexity of the debugger, in terms of efforts required to program it should not be a prohibiting factor either. A debugger is used, as often as, if not more than the compiler itself. Conventional Debuggers[2] which call themselves 'High Level Debugger' are no doubt better than 'Dumpers' but are still too primitive to help user in any substantial manner. Most of them are confined to setting up of breakpoints and examination of unstructured variables. SHD strives to match its level of commands with the thought process of the user for maximum efficiency in detecting bugs.