Polynomial, Neural Network, and Spline Wavelet Models for Continuous Wavelet Transform of Signals

In this paper a modified wavelet synthesis algorithm for continuous wavelet transform is proposed, allowing one to obtain a guaranteed approximation of the maternal wavelet to the sample of the analyzed signal (overlap match) and, at the same time, a formalized representation of the wavelet. What distinguishes this method from similar ones? During the procedure of wavelets’ synthesis for continuous wavelet transform it is proposed to use splines and artificial neural networks. The paper also suggests a comparative analysis of polynomial, neural network, and wavelet spline models. It also deals with feasibility of using these models in the synthesis of wavelets during such studies like fine structure of signals, as well as in analysis of large parts of signals whose shape is variable. A number of studies have shown that during the wavelets’ synthesis, the use of artificial neural networks (based on radial basis functions) and cubic splines enables the possibility of obtaining guaranteed accuracy in approaching the maternal wavelet to the signal’s sample (with no approximation error). It also allows for its formalized representation, which is especially important during software implementation of the algorithm for calculating the continuous conversions at digital signal processors and microcontrollers. This paper demonstrates the possibility of using synthesized wavelet, obtained based on polynomial, neural network, and spline models, during the performance of an inverse continuous wavelet transform.

Staying Tuned or Tuning Out? A Longitudinal Analysis of News-Avoiders on the Micro and Macro-Level

European media landscapes have changed into high-choice environments. Additionally, research suggests an increasing number of people are staying away from the news (news-avoiders). In two studies, we approach the topic on the macro and micro-level. In Study 1, we analyze on the macro-level how news-avoiders in 18 EU countries have developed since the late 1980s and what role country variables play in this context. We use multilevel spline models to distinguish between two phases of technological advancement. Findings suggest an increase in news-avoiders in the phase when the Internet was available. The market share of public-service broadcasters is no discerning factor for the number of news-avoiders in different countries. Using German panel-data, Study 2 takes a closer look at the news-avoiders in the Internet phase. We also show the increase of news-avoiders on the micro-level, but do not identify an increasing impact of political interest on news avoidance over time.

High and Low Haemoglobin Levels in Early Pregnancy Are Associated to a Higher Risk of Miscarriage: A Population-Based Cohort Study

To evaluate whether women with anaemia or high haemoglobin (Hb) in early pregnancy would be at higher risk of miscarriage, this population-based cohort study involved 9453 women whose pregnancies were monitored at primary care centres between 2007 and 2012. The computerised clinical histories were used to collect: Hb measurements (up to 14 weeks of gestation), miscarriage before or by 24 weeks of gestation, and other maternal characteristics. The relation between anaemia (Hb < 110 g/L), normal Hb (110–140 g/L, reference), and high Hb concentrations (≥140 g/L) with miscarriage were expressed as adjusted OR with 95%CI. Restricted cubic spline models were applied to evaluate the dose-response relationships. A total of 520 (5.5%) women were recorded as having a miscarriage. The rate of miscarriage in anaemia, normal Hb, and high Hb concentrations was 8.4%, 5.1%, and 10.2%, respectively. Compared with women with normal Hb at the first trimester, the multivariable-adjusted OR for miscarriage was 2.11 (95%CI, 1.38–3.21) for women with anaemia and 1.83 (95%CI, 1.29–2.58) for women with high Hb. Hb concentrations showed a U-shaped association with miscarriage, with the lowest incidence among women with Hb of 120–130 g/L. These data highlight the importance of considering anaemia and high Hb levels in early pregnancy as harmful indicators for miscarriage.

Effects of exposure to low-dose ionizing radiation on changing platelets: a prospective cohort study

Background Numerous studies have concentrated on high-dose radiation exposed accidentally or through therapy, and few involve low-dose occupational exposure, to investigate the correlation between low-dose ionizing radiation and changing hematological parameters among medical workers. Methods Using a prospective cohort study design, we collected health examination reports and personal dose monitoring data from medical workers and used Poisson regression and restricted cubic spline models to assess the correlation between changing hematological parameters and cumulative radiation dose and determine the dose-response relationship. Results We observed that changing platelet of 1265 medical workers followed up was statistically different among the cumulative dose groups (P = 0.010). Although the linear trend tested was not statistically significant (Ptrend = 0.258), the non-linear trend tested was statistically significant (Pnon-linear = 0.007). Overall, there was a correlation between changing platelets and cumulative radiation dose (a change of βa 0.008 × 109/L during biennially after adjusting for gender, age at baseline, service at baseline, occupation, medical level, and smoking habits; 95% confidence interval [CI] = 0.003,0.014 × 109/L). Moreover, we also found positive first and then negative dose-response relationships between cumulative radiation dose and changing platelets by restricted cubic spline models, while there were negative patterns of the baseline service not less than 10 years (− 0.015 × 109/L, 95% CI = − 0.024, − 0.007 × 109/L) and radiation nurses(− 0.033 × 109/L, 95% CI = − 0.049, − 0.016 × 109/L). Conclusion We concluded that although the exposure dose was below the limit, medical workers exposed to low-dose ionizing radiation for a short period of time might have increased first and then decreased platelets, and there was a dose-response relationship between the cumulative radiation dose and platelets changing.

CYBER CRIMES, FINANCIAL OFFENSES AND LEGALIZATION OF CRIMINAL INCOME IN THE CONDITIONS OF DIGITALIZATION OF UKRAINE'S ECONOMY

Digitalization of the economy is identified as one of the leading and priority fields of Ukraine's development. The complexity and development of new information technologies leads to the development of new methods and approaches to economic and financial activities. But despite the significant useful effects of digitalization, preferable conditions are also created for criminals. It is becoming easier to launder money, commit fraud with financial instruments and create new opportunities for cybercriminals. These processes lead to a detailed analysis of the determinants of criminal activity that has the greatest impact on the economy. The simulation identifies trends in the relationship of cybercrime, financial offenses and money laundering with generalized characteristics of fintech development, the number of reports of suspicious transactions submitted to the SCF and the level of development of key areas of financial activity: banks, insurance companies and stock exchanges. The study identified the direction and extent of the impact of digitalization, financial monitoring and activities of banking institutions, stock exchanges, insurance companies on the number of financial and cybercrimes and money laundering. MAR-spline models of interdependence between regressors and regressions were built. The adequacy of the constructed models is checked and confirmed. The advantage of MAR-spline models determines the possibility of calculating the limit values of factor features. The presence of the influence of factor features is traced either only until the achievement of certain threshold values or only when these threshold values are exceeded. Some factor characteristics had a multiplier effect on the resulting offenses. The practical application of the FinTech spline model of the interdependence of innovation and financial and cybercrime through financial institutions will benefit both financial intermediaries and users of the financial system, as well as government regulators and supervisors. Such a model can be useful and interesting to international organizations, investors and developers of regulatory standards, banking institutions, and other scientists conducting research in this area.

Extrapolation of Survival Curves Using Standard Parametric Models and Flexible Parametric Spline Models: Comparisons in Large Registry Cohorts with Advanced Cancer

Background It is often important to extrapolate survival estimates beyond the limited follow-up times of clinical trials. Extrapolated survival estimates can be highly sensitive to model choice; thus, appropriate model selection is crucial. Flexible parametric spline models have been suggested as an alternative to standard parametric models; however, their ability to extrapolate is not well understood. Aim To determine how well standard parametric and flexible parametric spline models predict survival when fitted to registry cohorts with artificially right-censored follow-up times. Methods Adults with advanced breast, colorectal, small cell lung, non–small cell lung, or pancreatic cancer with a potential follow-up time of 10 y were selected from the SEER 1973–2015 registry data set. Patients were classified into 15 cohorts by cancer and age group at diagnosis (18–59, 60–69, 70+ y). Follow-up times for each cohort were right censored at 20%, 35%, and 50% survival. Standard parametric models (exponential, Weibull, Gompertz, log-logistic, log-normal, generalized gamma) and spline models (proportional hazards, proportional odds, normal/probit) were fitted to the 10-y data set and the 3 right-censored data sets. Predicted 10-y restricted mean survival time and percentage surviving at 10 y were compared with the observed values. Results Across all data sets, the spline odds and spline normal models most frequently gave accurate predictions of 10-y survival outcomes. Visually, spline models tended to demonstrate better fit to the observed hazard functions than standard parametric models, both in the censored and 10-y data. Conclusions In these cohorts, where there was little uncertainty in the observed data, the spline models performed well when extrapolating beyond the observed data. Spline models should be routinely included in the set of models that are fitted when extrapolating cancer survival data.

Assessment of spline functions for estimating growth curve parameters of FUNAABAlpha chickens

The objective of this study was to fit four spline linear regression models to describe the growth of FUNAAB-Alpha Chickens (FAC). Body weight records of 300 FAC raised from day old till the 20th week were used to fit spline models of 3 (SP3), 4 (SP4), 5 (SP5) and 6 knots (SP6) using the REG procedure of SAS®. The data were first plotted to determine the most appropriate location of knots and they were placed at 4th, 10 th and 16 th week of age for SP3; 4th, 8th, 12th and 16th week for SP4; 4th, 7th, 10th, 14th and 18th week for SP5 and 3rd, 6th, 9th, 12th, 15 th and 18 th week for SP6, respectively. The hatch weight predicted by SP3 was observed to be highest while SP6 predicted the lowest hatch weight for male and female FAC. Regression coefficients ranged from -38.47 to 47.46 and -39.40 to 40.47 for the male and female, respectively. For all the models, the highest magnitude of these coefficients were estimated at early ages after hatching (at 3 to 10 weeks of age). Based on Bayesian Information Criterion (BIC) and Akaike Information Criterion (AIC) as the goodness-of-fit selection criteria, SP3 had the lowest value for AIC and BIC for male FAC while SP4 had the lowest value of AIC and BIC for the female FAC. It was concluded that spline models of lower knots (SP3 and SP4) were the best fit to describe the growth of male and female FAC respectively, and that growth rate at early stages of life of FAC may be good predictors of later growth performance.

The impact of time-updated resting heart rate on cause-specific mortality in a random middle-aged male population: a lifetime follow-up

Background High resting heart rate (RHR) is associated with increased adverse events. However, the long-term prognostic value in a general population is unclear. We aimed at investigating the impact of RHR, based on baseline as well as time-updated, on mortality in a middle-aged men cohort. Methods A random population sample of 852 men, all born in 1913 was followed from age 50 until age 98 with repeated examinations including RHR during 48 years. The impact of baseline and time-updated RHR on cause-specific mortality was assessed using Cox proportional hazards models and cubic spline models. Results Baseline RHR ≥90 beats per minute (bpm) was associated with higher all-cause mortality as compared to RHR 60–70 bpm (hazard ratio [HR] 1.60, 95% confidence interval [CI] 1.17–2.19, P=0.003), but not with cardiovascular (CV) mortality. A time-updated RHR &lt;60 bpm (HR 1.41, 95% CI 1.07–1.85, P=0.014) and a time-updated RHR of 70–80 bpm (HR 1.34, 95% CI 1.02–1.75, P=0.036) were both associated with higher CV mortality as compared with the RHR of 60–70 bpm after multivariable adjustment. Analyses using cubic spline models confirmed that the association of time-updated RHR with all-cause and CV mortality complied with a U-shaped curve with 60 bpm as a reference. Conclusions In this middle-aged men cohort, a time-updated RHR at 60–70 bpm was associated with lowest CV mortality, suggesting that time-updated RHR could be a useful long-term prognostic index in the general population. RHR and Mortality Funding Acknowledgement Type of funding source: Public grant(s) – National budget only. Main funding source(s): In recent years, funding has been received from the Swedish state under the agreement between the Swedish government and the county councils relating to the economic support of research and education under the ALF agreement (ALFGBG-721351).