Absolute Prediction Research Articles

Scope 3 emissions: Data quality and machine learning prediction accuracy

Investors’ sophistication on climate risk is increasing and as part of this they require high-quality and comprehensive Scope 3 emissions data. Accordingly, we investigate Scope 3 emissions data divergence (across different providers), composition (which Scope 3 categories are reported) and whether machine-learning models can be used to predict Scope 3 emissions for non-reporting firms. We find considerable divergence in the aggregated Scope 3 emissions values from three of the largest data providers (Bloomberg, Refinitiv Eikon, and ISS). The divergence is largest for ISS, as it replaces reported Scope 3 emissions with estimates from its economic input-output and life cycle assessment modelling. With respect to the composition of Scope 3 emissions, firms generally report incomplete composition, yet they are reporting more categories over time. There is a persistent contrast between relevance and completeness in the composition of Scope 3 emissions across sectors, with low materiality categories such as travel emissions being reported more frequently than typically high materiality ones, such as the use of products and processing of sold products. Finally, machine learning algorithms can improve the prediction accuracy of the aggregated Scope 3 emissions by up to 6% and up to 25% when each category is estimated individually and aggregated into total Scope 3 emissions. However, absolute prediction performance is low even with the best models, with the accuracy of estimates primarily limited by low observations in specific Scope 3 categories. We conclude that investors should be cognizant of Scope 3 emissions data divergence, incomplete reporting of Scope 3 categories, and that predictions for non-reporting firms have high absolute errors even when using machine learning models. For both reported and estimated data, caveat emptor applies.

An improved simplified model of self-piercing riveted joints for predicting quasi-static mechanical behavior of steel–aluminum hybrid components

The problem that the existing models could not accurately predict sheet deformation modes in the riveting area led to low prediction accuracy in crash simulations for steel-aluminum hybrid structural components in vehicle bodies. An improved simplified model with rigid and washer-processed elements (CONSTRAINED_SPR4_RIGWE) is proposed to calibrate the mechanical behavior of self-piercing riveted (SPR) joints based on the *CONSTRAINED_SPR4 model. The diameters of rigid and washer-processed elements in the upper and lower sheets were based on the diameters of the rivet and bottom of the SPR joint, respectively; the influence domain in the joint was determined by the diameter of the rivet cap. The parameters of the two models were calibrated using the experimental results of the lap shear, pull-out, mixed tension–shear, and peeling tests, as well as the parameters from the detailed three-dimensional model. Compared with the *CONSTRAINED_SPR4 model, the CONSTRAINED_SPR4_RIGWE model had a higher stiffness value for pull-out and peeling specimens, which was closer to the experimental result. T-shaped and C-shaped structural components of DP590 high-strength steel and AC43500-T7 cast aluminum sheets were prepared using SPR joints. For the T-R4 component, the absolute prediction error of the CONSTRAINED_SPR4_RIGWE model in terms of the energy absorption was lower than that of the *CONSTRAINED_SPR4 model, with the former being 2.48 % and the latter being 16.53 %. For the C-R8 component, the absolute maximum prediction errors of the CONSTRAINED_SPR4_RIGWE model in terms of peak force, failure displacement, and energy absorption increased from 16.82 % to 149.42 %, as compared with those of the *CONSTRAINED_SPR4 model.

High capacity reversible data hiding with contiguous space in encrypted images

Reversible data hiding in encrypted images (RDHEI) is used to embed secret data into an encrypted image to ensure data security and image privacy. However, a considerable portion of the existing methods do not achieve a high data embedding rate (ER) with complete reconstruction of the original image. Considering the aforementioned limitation, this paper proposes a new RDHEI method with contiguous space. The proposed method pre-processes the original image by dividing it into non-overlapping blocks to determine a label map based on absolute prediction errors. Next, the label map is encoded, and bit-plane shifting is done to create contiguous space for embedding. Thus, the proposed method effectively exploits local and overall image spatial correlation and creates a larger space for embedding in a bit-plane-wise manner. Compared to the state-of-the-art methods, the ER of the proposed method is significantly increased while ensuring complete reversibility and lossless extraction of the hidden data.

Refractive Predictability of a Swept Source Optical Coherence Tomography Biometer in Long and Short Eyes Implanted with Extended Depth of Focus Intraocular Lenses.

To determine the refractive predictability of Argos (Movu, a Santec company) measurements and the Barrett Universal II formula in long and short eyes implanted with an extended depth of focus (EDOF) intraocular lens (IOL). This retrospective, non-interventional study included 86 eyes (55 long and 31 short) of 55 patients. Preoperative biometry was performed using the Argos. Preoperative IOL power formulas were the preprogrammed Barrett Universal II (BUII). Data were collected for refractive outcomes, postoperative prediction error (directional and absolute), and monocular corrected distance visual acuity (CDVA, Snellen). The mean absolute prediction error for BUII was 0.27 ± 0.26 D overall, 0.24 ± 0.20 D in long eyes, and 0.33 ± 0.33 D in short eyes. Overall, the percentage of eyes with ≤ 0.5 D prediction error was 84% for BUII. In long eyes, the percentage of eyes with ≤ 0.5 D prediction error was 90% for BUII. In short eyes, the percentage of eyes with ≤ 0.5 D prediction error was 74% for BUII. The percentage of eyes with ≤ 0.5 D of MRSE was 89% for long eyes and 94% for short eyes. Visual acuities were excellent in both long and short eyes, with > 90% of eyes 20/25 or better in each group. The prediction error of Argos using BUII was low in long and short eyes at one month after EDOF IOL implantation.

Accuracy of the PEARL-DGS Formula for Intraocular Lens Power Calculation in Post–Myopic Laser Refractive Corneal Surgery Eyes

PurposeTo investigate the accuracy of the PEARL-DGS formula for intraocular lens (IOL) power calculation in post-myopic laser refractive corneal surgery eyes. DesignRetrospective case series MethodsA total of 139 eyes of 139 patients (mean axial length: 27.4 ± 2.1 mm), who had prior myopic laser refractive corneal surgery and subsequent cataract surgery using Tecnis ZCB00 from March 2018 to February 2023 were included. Refractive outcomes of five formulas (Barrett True K, Haigis-L, Hoffer-QST, PEARL-DGS, and Shammas-PL) were evaluated. Prediction error was defined as the difference between the measured and predicted postoperative refractive spherical equivalent using the IOL power actually implanted. Mean prediction error (MPE), median absolute prediction error (MedAE), mean absolute prediction error were calculated. ResultsWithout constant optimization, the PEARL-DGS resulted in a MPE of +0.05 ± 0.65 D, while the other formulas resulted in myopic shifts. The MedAEs of the formulas were 0.39, 0.53, 0.65, 0.85, and 1.11 D for the PEARL-DGS, Hoffer-QST, Barrett True K, Shammas-PL, and Haigis-L, respectively, in order of magnitude (P < 0.05). With constant optimization, there were no statistically significant differences in the MedAEs among the five formulas (P = 0.388). ConclusionsIn comparison to other IOL formulas, the PEARL-DGS resulted in better refractive outcomes after cataract surgery in post-myopic laser refractive corneal surgery eyes without constant optimization. We suggest that PEARL-DGS be considered as the first choice for IOL power calculation in these eyes when the clinicians do not have their optimized constants.

Approaches to Account for Colon Absorption in Physiologically Based Biopharmaceutics Modeling of Extended-Release Drug Products.

The rate and extent of colon absorption are important determinants of the in vivo performance of extended-release (ER) drug products. The ability to appropriately predict this at different stages of development using mechanistic physiologically based biopharmaceutic modeling (PBBM) is highly desirable. This investigation aimed to evaluate the prediction performance of three different approaches to account for colon absorption in predictions of the in vivo performance of ER drug product variants with different in vitro release profiles. This was done by mechanistic predictions of the absorption and plasma exposure of the ER drug products using GastroPlus and GI-Sim for five drugs with different degrees of colon absorption limitations in humans. Colon absorption was accounted for in the predictions using three different approaches: (1) by an a priori approach using the default colon models, (2) by fitting the colon absorption scaling factors to the observed plasma concentration-time profiles after direct administration to the colon in humans, or (3) from the ER drug product variant with the slowest in vitro release profile. The prediction performance was evaluated based on the percentage prediction error and the average absolute prediction error (AAPE). Two levels of acceptance criteria corresponding to highly accurate (AAPE ≤ 20%) and accurate (AAPE 20-50%) predictions were defined prior to the evaluation. For the a priori approach, the relative bioavailability (Frel), AUC0-t, and Cmax of the ER drug product variants for the low to medium colon absorption limitation risk drugs was accurately predicted with an AAPE range of 11-53 and 8-59% for GastroPlus and GI-Sim, respectively. However, the prediction performance was poor for the high colon absorption limitation risk drugs. Moreover, accounting for the human regional colon absorption data in the models did not improve the prediction performance. In contrast, using the colon absorption scaling factors derived from the slowest ER variant significantly improved the prediction performance regardless of colon absorption limitation, with a majority of the predictions meeting the high accuracy criteria. For the slowest ER approach, the AAPE ranges were 5-24 and 5-32% for GastroPlus and GI-Sim, respectively, excluding the low permeability drug. In conclusion, the a priori PBBM can be used during candidate selection and early product design to predict the in vivo performance of ER drug products for low to medium colon absorption limitation risk drugs with sufficient accuracy. The results also indicate a limited value in performing human regional absorption studies in which the drug is administered to the colon as a bolus to support PBBM development for ER drug products. Instead, by performing an early streamlined relative bioavailability study with the slowest relevant ER in vitro release profile, a highly accurate PBBM suitable for ER predictions for commercial and regulatory applications can be developed, except for permeability-limited drugs.

Predicting absolute aqueous solubility by applying a machine learning model for an artificially liquid-state as proxy for the solid-state.

In this study, we use machine learning algorithms with QM-derived COSMO-RS descriptors, along with Morgan fingerprints, to predict the absolute solubility of drug-like compounds. The QM-derived descriptors account for the molecular properties of the solute, i.e., the solute-solute interactions in an artificial-liquid-state (super-cooled liquid), and the solute-solvent interactions in solution. We employ two main approaches to predict solubility: (i) a hypothetical pathway that involves melting the solute at room temperature T = T¯ ([Formula: see text]) and mixing the artificially liquid solute into the solvent ([Formula: see text]). In this approach [Formula: see text] is predicted using machine learning models, and the [Formula: see text] is obtained from COSMO-RS calculations; (ii) direct solubility prediction using machine learning algorithms. The models were trained on a large number of Bayer in-house compounds for which water solubility data is available at physiological pH of 6.5 and ambient temperature. We also evaluated our models using external datasets from a solubility challenge. Our models present great improvements compared to the absolute solubility prediction with the QSAR model for the artificial liquid state as implemented in the COSMOtherm software, for both in-house and external datasets. We are furthermore able to demonstrate the superiority of QM-derived descriptors compared to cheminformatics descriptors. We finally present low-cost alternative models using fragment-based COSMOquick calculations with only marginal reduction in the quality of predicted solubility.

Contactless estimation of apple fruit respiration rate using machine learning models based on laser speckle imaging

This article presents the application of the laser speckle imaging method, a nondestructive and contactless method monitoring the gas exchange rate of intact apples. Red and infrared lasers in a typical setup for laser speckle imaging were used. A new parameter, speckle pattern relaxation time, has been proposed to evaluate speckle dynamics. Machine learning methods were used to develop a set of predictive models calibrated and validated against the respiration rate of two apple fruit cultivars measured with the flush system. The model with the highest performance used three variables: the speckle pattern relaxation time (τRED or τIR), fruit mass, and categorical variables describing apple varieties. This model provided satisfactorily low values of mean absolute prediction errors of 6.04%. Data from laser light scattering measurements combined with modern machine learning algorithms provided a nondestructive and fast method for estimating the apple fruit respiration rate. The developed solution has the potential for a wide range of industrial applications, especially in fruit storage, where the fruit respiration rate indicates optimal storage conditions.

Assessing the predictive performance of creep models using absolute rather than squared prediction errors: an application to 2.25Cr-1Mo steel and 316H stainless steel

ABSTRACT A reliable means of assessing the accuracy of a creep model’s predictions is fundamental to safe power plant operation. This paper introduces a method of decomposing the mean absolute prediction error for such a purpose to overcome the limitations that are inherent in the traditional approach of squaring prediction errors to prevent over and underestimates of life offsetting each other. When this method is applied to 2.25Cr-1Mo steel and 316 H stainless steel, it was found that squared errors leads to overestimates of the average prediction error associated with a particular creep model, and it also dramatically underestimates the proportion of this error that is systematic in nature. These differences were more noticeable for 316 H stainless steel.

Accuracy of Different Lens Power Calculation Formulas in Patients With Mature Cataracts.

Introduction To compare the prediction accuracy of lens power calculation formulas in patients undergoing mature cataract surgery. Methods A total of 90 operations involving the Alcon SA60AT IOL implant (Alcon, Geneva, Switzerland) were analyzed in terms of mean refractive prediction error (PE) and mean absolute prediction error (MAE) using backward calculation in a retrospective design. Results A negative PE was observed in SRK/T, Holladay 1, Holladay 2, Hoffer Q, Haigis, and Emmetropia Verifying Optical (EVO) formulas. In contrast, positive PEs were observed in Barrett Universal II (BAUII), Kane, and Radial Basis Function (RBF) formulas. Negative PE was observed with all formulas, except BAUII, in patients with a shallow anterior chamber depth (ACD). While the SRK/T, Holladay 1, BAU, Kane, and RBF formulas demonstrated positive PE, the Holladay 2, Hoffer Q, Haigis, and EVO formulas indicated negative PE. In patients with deep ACD, positive PE was observed in all formulas, barring Holladay 2 and EVO. No significant differences were identified between the formulas concerning MAE and percentages of 0.25 diopter (D), 0.50 D, 0.75 D, and 1.0 D across all study groups. Conclusion Although the new generation formulas provide very good results, achieving the best with a single formula is still impossible.

Modeling the protein binding non-linearity in population pharmacokinetic model of valproic acid in children with epilepsy: a systematic evaluation study.

Background: Several studies have investigated the population pharmacokinetics (popPK) of valproic acid (VPA) in children with epilepsy. However, the predictive performance of these models in the extrapolation to other clinical environments has not been studied. Hence, this study evaluated the predictive abilities of pediatric popPK models of VPA and identified the potential effects of protein binding modeling strategies. Methods: A dataset of 255 trough concentrations in 202 children with epilepsy was analyzed to assess the predictive performance of qualified models, following literature review. The evaluation of external predictive ability was conducted by prediction- and simulation-based diagnostics as well as Bayesian forecasting. Furthermore, five popPK models with different protein binding modeling strategies were developed to investigate the discrepancy among the one-binding site model, Langmuir equation, dose-dependent maximum effect model, linear non-saturable binding equation and the simple exponent model on model predictive ability. Results: Ten popPK models were identified in the literature. Co-medication, body weight, daily dose, and age were the four most commonly involved covariates influencing VPA clearance. The model proposed by Serrano et al. showed the best performance with a median prediction error (MDPE) of 1.40%, median absolute prediction error (MAPE) of 17.38%, and percentages of PE within 20% (F20, 55.69%) and 30% (F30, 76.47%). However, all models performed inadequately in terms of the simulation-based normalized prediction distribution error, indicating unsatisfactory normality. Bayesian forecasting enhanced predictive performance, as prior observations were available. More prior observations are needed for model predictability to reach a stable state. The linear non-saturable binding equation had a higher predictive value than other protein binding models. Conclusion: The predictive abilities of most popPK models of VPA in children with epilepsy were unsatisfactory. The linear non-saturable binding equation is more suitable for modeling non-linearity. Moreover, Bayesian forecasting with prior observations improved model fitness.

Accuracy of Seven Modern Online IOL Formulas in Eyes With Axial Lengths Longer Than 30 mm.

To evaluate the accuracy of newer online intraocular lens (IOL) formulas in extremely elongated eyes (axial length > 30 mm). This retrospective case series study included 236 patients (236 eyes). Postoperative refractive outcomes of the Barrett Universal II (BU II), Cooke K6 (K6), Emmetropia Verifying Optical (EVO) 2.0, Hoffer QST (HQST), Kane, Pearl-DGS, and Radial Basis Function (RBF) 3.0 formulas were compared. Subgroup analysis was performed in the extreme myopia group 1 (30 < axial length ≤ 32 mm), extreme myopia group 2 (32 < axial length ≤ 35 mm), and meniscus IOL group. The root mean square absolute prediction error (RMSAE) and proportions of eyes of prediction errors within ±0.50 diopters (D) were calculated for statistical analysis. For the extreme myopia group 1, RBF 3.0 achieved the lowest RMSAE (0.361) and EVO 2.0 showed the highest proportion of eyes within ±0.50 diopters (85.06%). For the extreme myopia group 2, the RMSAE of the K6 (0.442) and EVO 2.0 (0.475) was significantly lower than the BU II (0.610), Kane (0.641), and HQST (0.759, P ≤ .016) formulas. In the meniscus IOL group, the K6 formula showed the lowest RMSAE (0.402) and the highest percentage within ±0.50 diopters (84.31%). The EVO 2.0 and K6 formulas are recommended for IOL power calculation in eyes with extreme myopia. Modern artificial intelligence-based formulas should be used cautiously when the axial length is longer than 32 mm or meniscus IOLs are implanted. [J Refract Surg. 2023;39(10):705-710.].

Online learning of smooth functions

In this paper, we study the online learning of real-valued functions where the hidden function is known to have certain smoothness properties. Specifically, for q≥1, let Fq be the class of absolutely continuous functions f:[0,1]→R such that ‖f′‖q≤1. For q≥1 and d∈Z+, let Fq,d be the class of functions f:[0,1]d→R such that any function g:[0,1]→R formed by fixing all but one parameter of f is in Fq. For any class of real-valued functions F and p>0, let optp(F) be the best upper bound on the sum of pth powers of absolute prediction errors that a learner can guarantee in the worst case. In the single-variable setup, we find new bounds for optp(Fq) that are sharp up to a constant factor. We show for all ε∈(0,1) that opt1+ε(F∞)=Θ(ε−12) and opt1+ε(Fq)=Θ(ε−12) for all q≥2. We also show for ε∈(0,1) that opt2(F1+ε)=Θ(ε−1). In addition, we obtain new exact results by proving that optp(Fq)=1 for q∈(1,2) and p≥2+1q−1. In the multi-variable setup, we establish inequalities relating optp(Fq,d) to optp(Fq) and show that optp(F∞,d) is infinite when p<d and finite when p>d. We also obtain sharp bounds on learning F∞,d for p<d when the number of trials is bounded.

Predictive Modeling of Signal Degradation in Urban VANETs Using Artificial Neural Networks

In urban Vehicular Ad Hoc Network (VANET) environments, buildings play a crucial role as they can act as obstacles that attenuate the transmission signal between vehicles. Such obstacles lead to multipath effects, which could substantially impact data transmission due to fading. Therefore, quantifying the impact of buildings on transmission quality is a key parameter of the propagation model, especially in critical scenarios involving emergency vehicles where reliable communication is of utmost importance. In this research, we propose a supervised learning approach based on Artificial Neural Networks (ANNs) to develop a predictive model capable of estimating the level of signal degradation, represented by the Bit Error Rate (BER), based on the obstacles perceived by moving emergency vehicles. By establishing a relationship between the level of signal degradation and the encountered obstacles, our proposed mechanism enables efficient routing decisions being made prior to the transmission process. Consequently, data packets are routed through paths that exhibit the lowest BER. To collect the training data, we employed Network Simulator 3 (NS-3) in conjunction with the Simulation of Urban MObility (SUMO) simulator, leveraging real-world data sourced from the OpenStreetMap (OSM) geographic database. OSM enabled us to gather geospatial data related to the Two-Dimensional (2D) geometric structure of buildings, which served as input for our Artificial Neural Network (ANN). To determine the most suitable algorithm for our ANN, we assessed the accuracy of ten learning algorithms in MATLAB, utilizing five key metrics: Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Correlation Coefficient (R), and Maximum Prediction Error (MaxPE). For each algorithm, we conducted fifteen iterations based on ten hidden neurons and gauged its accuracy against the aforementioned metrics. Our analysis highlighted that the ANN underpinned by the Conjugate Gradient With Powell/Beale Restarts (CGB) learning algorithm exhibited superior performance in terms of MSE, RMSE, MAE, R, and MaxPE compared to other algorithms such as Levenberg–Marquardt (LM), Bayesian Regularization (BR), BFGS Quasi-Newton (BFG), Resilient Backpropagation (RP), Scaled Conjugate Gradient (SCG), Fletcher–Powell Conjugate Gradient (CGF), Polak–Ribiére Conjugate Gradient (CGP), One-Step Secant (OSS), and Variable Learning Rate Backpropagation (GDX). The BER prediction by our ANN incorporates the TWO-RAY Ground (TRG) propagation model, an adjustable parameter within NS-3. When subjected to 300 new samples, the trained ANN’s simulation outcomes illustrated its capability to learn, generalize, and successfully predict the BER for a new data instance. Overall, our research contributes to enhancing the performance and reliability of communication in urban VANET environments, especially in critical scenarios involving emergency vehicles, by leveraging supervised learning and artificial neural networks to predict signal degradation levels and optimize routing decisions accordingly.

Medium-term visual and refractive outcomes in patients with pseudoexfoliation implanted with a toric or multifocal intraocular lens.

To report medium-term visual acuity and refractive outcomes of patients with pseudoexfoliation implanted with toric or multifocal intraocular lenses (IOLs). We retrospectively reviewed patients with pseudoexfoliation who had undergone phacoemulsification between 2016 and 2020 with at least 24 months follow-up. Mean follow-up was 44.17 ± 14.95 months. Toric IOLs were implanted in 48 eyes: mean uncorrected LogMAR visual acuity one month after surgery was 0.03 ± 0.09, decreasing to 0.08 ± 0.11 at the last visit (p = .01). Mean refractive cylinder was -0.12 ± 0.36 diopters one month after surgery and -0.25 ± 0.44 diopters at the last follow-up visit, p = .012. Multifocal IOLs were implanted in 42 patients: binocular uncorrected distance visual acuity was -0.02 ± 0.04 one month after surgery and 0.01 ± 0.05 at the last visit (p = .004); near acuity was 0.01 ± 0.03 and 0.04 ± 0.06 respectively (p = .001). In eyes with pseudoexfoliation, absolute spherical equivalent prediction error was 0.22 ± 0.20 diopters for toric and 0.21 ± 0.19 diopters for multifocal IOLs. One month after surgery 68.6% of eyes with toric IOLs and 74.2% of eyes with multifocal IOLs were within ±0.25 diopters of target spherical equivalent and 91.6% and 90.5% were within ±0.5 diopters, respectively. Spherical equivalent did not change significantly during follow-up for either group. Prediction error in eyes with pseudoexfoliation implanted with toric or multifocal IOLs was low and similar to values reported for normal eyes. Postoperative refractive cylinder with toric IOLs was low, with little change during follow-up. Visual function in patients receiving multifocal IOLs was excellent. Therefore, the implantation of these IOLs in eyes with pseudoexfoliation does not seem to cause medium-term problems.

Ability of three dairy feed evaluation systems to predict postruminal outflows of nitrogenous compounds in dairy cows: A meta-analysis

Adequate prediction of postruminal outflow of protein fractions is the starting point for the determination of metabolizable protein supply in dairy cows. The objective of this meta-analysis was to compare the performance of 3 dairy feed evaluation systems (National Research Council [NRC], Cornell Net Protein and Carbohydrate System [CNCPS], and National Academies of Sciences, Engineering and Medicine [NASEM]) to predict outflows (g/d) of nonammonia nitrogren (NAN), microbial N (MiN), and nonammonia nonmicrobial N (NANMN). Predictions of rumen degradabilities (% of nutrient) of protein (RDP), NDF, and starch were also evaluated. The data set included 1,294 treatment means from 312 digesta flow studies. The 3 feed evaluation systems were compared using the concordance correlation coefficient (CCC), the ratio of root mean square prediction error (RMSPE) on standard deviation of observed values (RSR), and the slope between observed and predicted values. Mean and linear biases were deemed biologically relevant and are discussed if higher than a threshold of 5% of the mean of observed values. The comparisons were done on observed values adjusted or not for the study effect; the adjustment had a small effect on the mean bias but the linear bias reflected a response to a dietary change rather than absolute predictions. For the absolute predictions of NAN and MiN, CNCPS had the best-fit statistics (8% greater CCC; 6% lower RMSPE) without any bias; NRC and NASEM underpredicted NAN and MiN, and NASEM had an additional linear bias indicating that the underprediction of MiN increased at increased predictions. For NANMN, fit statistics were similar among the 3 feed evaluation systems with no mean bias; however, the linear bias with NRC and CNCPS indicated underprediction at low predictions and overprediction at elevated predictions. On average, the CCC were smaller and RSR ratios were greater for MiN versus NAN indicating increased prediction errors for MiN. For NAN responses to a dietary change, CNCPS also had the best predictions, although the mean bias with NASEM was not biologically relevant and the 3 feed evaluation systems did not present a linear bias. However, CNCPS, but not the 2 other feed evaluation systems, presented a linear bias for MiN, with responses being overpredicted at increased predictions. For NANMN, responses were overpredicted at increased predictions for the 3 feed evaluation systems, but to a lesser extent with NASEM. The site of sampling had an effect on the mean bias of MiN and NANMN in the 3 feed evaluation systems. The mean bias of MiN was higher in omasal than duodenal studies in the 3 feed evaluation systems (from 55 to 61 g/d) and this mean bias was twice as large when 15N labeling was used as a microbial marker compared with purines. Such a difference was not observed for duodenal studies. The reasons underlying these systematic differences are not clear as the type of measurements used in the current meta-analysis does not allow to delineate if one site or one microbial marker is yielding the "true" postruminal N outflows. Rumen degradabilities of protein was underpredicted with CNCPS, and RDP responses to a dietary change was underpredicted by the 3 feed evaluation systems with increased RDP predictions. Rumen degradability of NDF was underpredicted and had poor fit statistics for NASEM compared with CNCPS. Fit statistics were similar between CNCPS and NASEM for rumen degradability of starch, but with an underprediction of the response with NASEM and absolute values being overpredicted with CNCPS. Multivariate regression analyses showed that diet characteristics were correlated with prediction errors of N outflows in each feed evaluation system. Globally, compared with NAN and NANMN, residuals of MiN were correlated with several moderators in the 3 feed evaluation systems reflecting the complexity to measure and model this outflow. In addition, residuals of NANMN were correlated positively with RDP suggesting an overestimation of this parameter. In conclusion, although progress is still to be made to improve equations predicting postruminal N outflows, the current feed evaluation systems provide sufficient precision and accuracy to predict postruminal outflows of N fractions.

Dynamic mode decomposition and bivariate autoregressive short-term prediction of Earth rotation parameters

Abstract In this contribution two new approaches are applied to predict polar motion and length-of-day. The first one is based on Dynamic Mode Decomposition (DMD), that is purely data-driven and is capable of reconstructing and forecasting time series in one numerical procedure. The other one is based on a vector autoregression of order p – VAR(p), which is a vector counterpart of AR(p) that accounts for an evolution of variables in time and a coevolution with other variables. DMD was applied to polar motion and length-of-day whilst VAR(p) to a joint prediction of polar motion. A prediction experiment concerned 30-day forecast horizon with a 7-day shift. It was performed separately for years 2017–2022 giving 48 predictions within each year. This study uses IERS EOP 14 C04 (IAU2000) as a reference for all computations and a mean absolute prediction error (MAPE) as a measure of prediction quality. For DMD, MAPEs for x coordinate of the pole vary from 0.22–0.30 mas for the 1st day and 6.64–8.56 mas for the 30th day of prediction depending on the year whilst those values vary from 0.20–0.27 mas and 5.27–7.66 mas for VAR(p) based prediction. Corresponding values for y coordinate of the pole vary from 0.15–0.23 mas and 4.27–5.93 mas for DMD, whilst 0.13–0.21 mas and 3.46–3.82 mas for VAR(p). In case of LOD forecast, MAPEs vary from 0.023–0.031 ms for the 1st day and 0.142–0.205 ms for the 30th day depending on the year.

Correlation Between Early Trends of a Prognostic Biomarker and Overall Survival in Non-Small-Cell Lung Cancer Clinical Trials.

Overall survival (OS) is the primary end point in phase III oncology trials. Given low success rates, surrogate end points, such as progression-free survival or objective response rate, are used in early go/no-go decision making. Here, we investigate whether early trends of OS prognostic biomarkers, such as the ROPRO and DeepROPRO, can also be used for this purpose. Using real-world data, we emulated a series of 12 advanced non-small-cell lung cancer (aNSCLC) clinical trials, originally conducted by six different sponsors and evaluated four different mechanisms, in a total of 19,920 individuals. We evaluated early trends (until 6 months) of the OS biomarker alongside early OS within the joint model (JM) framework. Study-level estimates of early OS and ROPRO trends were correlated against the actual final OS hazard ratios (HRs). We observed a strong correlation between the JM estimates and final OS HR at 3 months (adjusted = 0.88) and at 6 months (adjusted = 0.85). In the leave-one-out analysis, there was a low overall prediction error of the OS HR at both 3 months (root-mean-square error [RMSE] = 0.11) and 6 months (RMSE = 0.12). In addition, at 3 months, the absolute prediction error of the OS HR was lower than 0.05 for three trials. We describe a pipeline to predict trial OS HRs using emulated aNSCLC studies and their early OS and OS biomarker trends. The method has the potential to accelerate and improve decision making in drug development.

Factors Associated with Refractive Prediction Error after Phacotrabeculectomy.

To compare refractive prediction errors between phacotrabeculectomy and phacoemulsification. Refractive prediction error was defined as the difference in spherical equivalent between the predicted value using the Barrett Universal II formula and the actual value obtained at postoperative one month. Forty-eight eyes that had undergone phacotrabeculectomy (19 eyes, open-angle glaucoma; 29 eyes, angle-closure glaucoma) were matched with 48 eyes that had undergone phacoemulsification by age, average keratometry value and axial length (AL), and their prediction errors were compared. The factors associated with prediction errors were analyzed by multivariable regression analyses. The phacotrabeculectomy group showed a larger absolute prediction error than the phacoemulsification group (0.51 ± 0.37 Diopters vs. 0.38 ± 0.22 Diopters, p = 0.033). Larger absolute prediction error was associated with longer AL (p = 0.010) and higher intraocular pressure (IOP) difference (p = 0.012). Hyperopic shift (prediction error > 0) was associated with shallower preoperative anterior chamber depth (ACD) (p = 0.024) and larger IOP difference (p = 0.031). In the phacotrabeculectomy group, the prediction error was inversely correlated with AL: long eyes showed myopic shift and short eyes hyperopic shift (p = 0.002). Surgeons should be aware of the possibility of worse refractive outcomes when planning phacotrabeculectomy, especially in eyes with high preoperative IOP, shallow ACD, and/or extreme AL.

Actual anterior–posterior corneal radius ratio in eyes with prior myopic laser vision correction according to axial length

We retrospectively evaluate the actual anterior–posterior (AP) corneal radius ratio in eyes with previous laser correction for myopia (M-LVC) according to axial length (AL) using biometry data exported from swept-source optical coherence tomography between January 2018 and October 2021 in a tertiary hospital (1018 eyes with a history of M-LVC and 19,841 control eyes). The AP ratio was significantly higher in the LVC group than in the control group. Further, it was significantly positively correlated with AL in the LVC group. We also investigated the impact of the AP ratio, AL and keratometry (K) on the absolute prediction error (APE) in 39 eyes that underwent cataract surgery after M-LVC. In linear regression analyses, there were significant correlations between APE and AL/TK, while APE and AP ratio had no correlation. The APE was significantly lower in the Barrett True-K with total keratometry (Barrett True-TK) than in the Haigis-L formula on eyes with AL above 26 mm and K between 38 and 40 D. In conclusion, in eyes with previous M-LVC, AP ratio increases with AL. The Barrett True-K or Barrett True-TK formulas are recommended rather than Haigis-L formula in M-LVC eyes with AL above 26 mm and K between 38 and 40D.