Original Prediction Method Research Articles

Practical application of the CCB model in Czechia

This research aimed to present a new bankruptcy prediction model and apply this original prediction method in practice. The Come Clean Bankruptcy (or CCB) model uses relevant financial indicators and ratios to detect the signs of impending financial distress in time so that the management can take appropriate measures to avoid it. The model was applied to the data reported by 199 entities operating in the textile/clothing industry in the Czech Republic. Analyzing data reported for the previous seven years enabled us to predict which companies are more likely to end in a difficult financial situation. Afterward, comparing these predictions with the actual development of those companies in 2013-2020 serves to verify the efficacy and usability of the model to corporate reality. The research has shown that companies that went bankrupt in the analyzed period represented only a fraction of the data set (roughly 4.5%). Despite the small number of financial failures occurring during the analyzed period, the CCB model could detect impending bankruptcy in one-third of the cases.

Adaptive time offset compensation for TDMA industrial wireless networks

Wireless network nodes working in industrial environment or mounted on mobile machines are exposed to frequent and unpredictable temperature changes, which impact local clock accuracy resulting in synchronization problems (time offset). These issues have to be taken into account when optimizing the network to avoid higher energy consumption and traffic congestion. The time offset problem can be alleviated using temperature driven compensation. In the paper we propose an efficient algorithm to resolve this problem taking into account slow and rapid temperature fluctuations. It combines the physical model of the node oscillator with an original prediction method based on the linear regression. It can be embedded in low cost microcontrollers with limited resources. Our approach needed some experimental studies of oscillator features and microcontroller capabilities. The implemented prototype nodes and performed tests proved high clock resilience for industrial environment as compared with other approaches.

Application of improved grey model residual modified fusion algorithm in dissolved gas forecasting

The power transformer directly affects whether the power system can operate safely and reliably, and the prediction of the dissolved gas content in the transformer oil is an effective means to discover transformer defects and latent faults, which helps to better carry out state maintenance work. This study extracts the time series data in transformer oil and analyses the interactions and effects between various gas components in the oil. Then, combined with the changing trend of the content of each component, the original sequence was processed using an accumulated method to extract the intrinsic characteristics of the sequence, thus establishing the MGM(1,n)(grey multivariate model). As the grey prediction model can reduce the prediction accuracy of the sequence with longer time span, this study combines the adaptive regression and Markov correction models to optimise the grey prediction model from the perspective of residual correction. Finally, the accuracy and effectiveness of the model are verified by comparing with the original prediction method.

FunFam protein families improve residue level molecular function prediction

BackgroundThe CATH database provides a hierarchical classification of protein domain structures including a sub-classification of superfamilies into functional families (FunFams). We analyzed the similarity of binding site annotations in these FunFams and incorporated FunFams into the prediction of protein binding residues.ResultsFunFam members agreed, on average, in 36.9 ± 0.6% of their binding residue annotations. This constituted a 6.7-fold increase over randomly grouped proteins and a 1.2-fold increase (1.1-fold on the same dataset) over proteins with the same enzymatic function (identical Enzyme Commission, EC, number). Mapping de novo binding residue prediction methods (BindPredict-CCS, BindPredict-CC) onto FunFam resulted in consensus predictions for those residues that were aligned and predicted alike (binding/non-binding) within a FunFam. This simple consensus increased the F1-score (for binding) 1.5-fold over the original prediction method. Variation of the threshold for how many proteins in the consensus prediction had to agree provided a convenient control of accuracy/precision and coverage/recall, e.g. reaching a precision as high as 60.8 ± 0.4% for a stringent threshold.ConclusionsThe FunFams outperformed even the carefully curated EC numbers in terms of agreement of binding site residues. Additionally, we assume that our proof-of-principle through the prediction of protein binding residues will be relevant for many other solutions profiting from FunFams to infer functional information at the residue level.

Displacement prediction of step-like landslide by applying a novel kernel extreme learning machine method

Landslide displacement prediction is an essential component for developing landslide early warning systems. In the Three Gorges Reservoir area (TGRA), landslides experience step-like deformations (i.e., periods of stability interrupted by abrupt accelerations) generally from April to September due to the influence of precipitation and reservoir scheduled level variations. With respect to many traditional machine learning techniques, two issues exist relative to displacement prediction, namely the random fluctuation of prediction results and inaccurate prediction when step-like deformations take place. In this study, a novel and original prediction method was proposed by combining the wavelet transform (WT) and particle swarm optimization-kernel extreme learning machine (PSO-KELM) methods, and by considering the landslide causal factors. A typical landslide with a step-like behavior, the Baishuihe landslide in TGRA, was taken as a case study. The cumulated total displacement was decomposed into trend displacement, periodic displacement (controlled by internal geological conditions and external triggering factors respectively), and noise. The displacement items were predicted separately by multi-factor PSO-KELM considering various causal factors, and the total displacement was obtained by summing them up. An accurate prediction was achieved by the proposed method, including the step-like deformation period. The performance of the proposed method was compared with that of the multi-factor extreme learning machine (ELM), support vector regression (SVR), backward propagation neural network (BPNN), and single-factor PSO-KELM. Results show that the PSO-KELM outperforms the other models, and the prediction accuracy can be improved by considering causal factors.

Improved BP neural network algorithm to wind power forecast

To constantly enhance the accuracy of wind power prediction and furthermore reduce the uncertainty of power grid dispatching, this study proposes an improved back propagation (BP) neural network algorithm. The original prediction method of BP neural network algorithm has been improved, and the traditional minimum square error (SE) perform function is abandoned. Maximum correntropy criteria (MCC) algorithm which is more conducive to deal with non-Gaussian error and big noise is introduced, and a new perform function is created. Through the analysis of examples, the feasibility of MCC algorithm is verified. Comparing to the traditional mean SE (MSE) perform function, MCC perform function could drop the limit error of prediction, reduce root MSE and increase the correlation between forecasting power and real power. The most important is that the prediction accuracy is enhanced.

Theoretical Screening of Novel Host Oxide Materials Suitable for Eu2+ Doped Yellow and Red Luminescence Phosphor

Theoretical screening of host materials suitable for Eu2+ doped phosphor applicable to light emitting diodes was carried by using computational chemistry techniques and crystal structure database. It is well known that an energy level of Eu 4f and 5d orbitals are shifted by the interaction with host crystals that causes a significant spread of Eu 5d orbital. Hence, a control of emission properties of phosphors would be possible by selecting a suitable host material. In this study, emission wavelength and efficiency for more than 8000 oxide structures, almost which are not well investigated as a host structure for Eu2+, were screened by our original prediction method. We reported that there is the relationship between crystal structure and emission colors of Eu2+ doped phosphors, and a quantitative structure-property relationship (QSPR) was applied to reveal it. Developed QSPR model is confirmed to predict the emission wavelength of phosphors we investigated. Our suggested method for the estimation of emission efficiency of Eu2+-doped phosphors is based on the structural information. We defined an alkaline-earth network (ANet) to represent the magnitude of the interaction of alkaline earth atoms with Eu2+ atoms in the host crystal. ANet shows an inversely relationship to an internal quantum efficiency, which supports that ANet is a good index for searching a high efficient Eu2+-doped phosphors qualitatively. It is noted that they are estimated only from the structural information of host materials. In this paper, the prediction methods of emission properties were applied to explore novel host structures for Eu2+ doped phosphors applicable for LED. For more than 8000 oxide crystal candidates for host structure, are screening. Several oxide structure suitable for Eu2+ doped phosphor for LED are suggested.

A hybrid approach to advancing quantitative prediction of tissue distribution of basic drugs in human

A general toxicity of basic drugs is related to phospholipidosis in tissues. Therefore, it is essential to predict the tissue distribution of basic drugs to facilitate an initial estimate of that toxicity. The objective of the present study was to further assess the original prediction method that consisted of using the binding to red blood cells measured in vitro for the unbound drug (RBCu) as a surrogate for tissue distribution, by correlating it to unbound tissue:plasma partition coefficients (Kpu) of several tissues, and finally to predict volume of distribution at steady-state (V ss) in humans under in vivo conditions. This correlation method demonstrated inaccurate predictions of V ss for particular basic drugs that did not follow the original correlation principle. Therefore, the novelty of this study is to provide clarity on the actual hypotheses to identify i) the impact of pharmacological mode of action on the generic correlation of RBCu–Kpu, ii) additional mechanisms of tissue distribution for the outlier drugs, iii) molecular features and properties that differentiate compounds as outliers in the original correlation analysis in order to facilitate its applicability domain alongside the properties already used so far, and finally iv) to present a novel and refined correlation method that is superior to what has been previously published for the prediction of human V ss of basic drugs. Applying a refined correlation method after identifying outliers would facilitate the prediction of more accurate distribution parameters as key inputs used in physiologically based pharmacokinetic (PBPK) and phospholipidosis models.

Structure-based prediction of the conductance properties of ion channels.

The HOLE procedure allows the prediction of the absolute conductance of an ion channel model from its structure. The original prediction method uses an empirically corrected Ohmic method. It is most successful, with predictions being reliable to within a factor of two. A new modification of the procedure is presented in which the self-diffusion coefficients of water molecules from molecular dynamics simulation are used to replace the empirical correction factor. A "prediction" of the conductance for the porin OmpF by the new method is made and shown to be very close to the experimental value. HOLE also allows the prediction of the effect that the addition of non-electrolyte polymers will have on channel conductance. The method has great potential to yield structural information from data provided by single channel recordings but needs further validation by making measurements on channels of known structure. Preliminary results are given of single channel records establishing the effects of non-electrolytes on the conductance of gramicidin D channels. As an example of the potential uses of the procedure application is made to examine the oligomerization of alpha-toxin (alpha-hemolysin) channels. A model for the alpha-toxin hexamer, based on the crystal structure for the heptamer, is generated using molecular mechanics methods. The compatibility of the structures with single channel conductance data is assessed using HOLE.