Traditional Prediction Methods Research Articles

Prediction of formation fracture pressure based on reinforcement learning and XGBoost

Abstract Clearly determining the magnitude of fracture pressure is a crucial indicator for fracturing design. Traditional methods for predicting fracture pressure suffer from challenges such as difficulties in obtaining required data, low prediction accuracy, and local limitations in application. In light of these issues, the article proposes a fracture pressure prediction model based on reinforcement learning and XGBoost utilizing geophysical well logging data. Based on the relevance analysis, optimal input parameters, including DEPTH, DEN, AC, GR, CRL, and RT, are selected from geophysical well logging data. We have developed a framework for a fracture pressure prediction model based on XGBoost, wherein hyperparameters are fine-tuned using an improved Q-learning algorithm. The optimized XGBoost model for fracture pressure prediction attains outstanding performance metrics, including an R 2 value of 0.992, a root mean square error of 0.006%, and a mean absolute error of 0.539%. In direct comparison with grid search, Bayesian optimization, and ant colony optimization, the improved Q-learning algorithm emerges as the most effective optimization approach. The predictions generated by the proposed method exhibit remarkable consistency with fracture pressure data measured on-site. This approach successfully addresses the shortcomings encountered with traditional fracture pressure prediction methods, such as inadequate accuracy, demanding data prerequisites, and constrained applicability.

Ontology-Based Landslide Knowledge Representation and Hazard Assessment Research

The occurrence of landslides is characterized by suddenness and unpredictability, with many complex relationships among various influencing factors. This study, by introducing the ontology of knowledge engineering, summarizes the characteristics of landslide knowledge and constructs a multidisciplinary, multi-level, and multi-perspective ontology of the landslide domain. Through in-depth analysis of existing landslide research systems and methods, the study defines the basic concepts, attributes, and their complex relationships and rules related to landslides. This research expresses landslide knowledge using the Web Ontology Language (OWL), enabling effective storage and retrieval of landslide knowledge, thus promoting the systematization and standardization of knowledge in the landslide domain. Furthermore, leveraging the high knowledge density feature of the constructed landslide domain ontology, this study extends traditional landslide evaluation and prediction methods by introducing more evaluation indicators based on the constructed landslide ontology. The comprehensive evaluation of landslide risk is conducted using the contribution-weighting method, providing new insights and tools for landslide risk management and decision-making.

The role of news sentiment in salmon price prediction using deep learning

This paper employs deep learning models and sentiment analysis to predict salmon spot prices. Our data includes historical price data and sentiment scores from 2018 to 2022. We extract sentiment scores from salmon-related news headlines by using FinBERT and TextBlob. We begin with price prediction using only historical price data and then introduce sentiment scores to improve the prediction accuracy of deep learning models. We find that the prediction performance of deep learning models outperforms traditional prediction methods in the salmon market. Our primary hybrid CNN-LSTM model outperforms other deep learning models and traditional models. Additionally, deep learning models incorporating sentiment scores exhibit reduced prediction errors. Our findings confirm the value of sentiment information in improving forecasting performance. These findings highlight the effectiveness and robustness of our CNN-LSTM model combined with sentiment analysis for price prediction in the salmon market.

Wfold: A new method for predicting RNA secondary structure with deep learning

Precise estimations of RNA secondary structures have the potential to reveal the various roles that non-coding RNAs play in regulating cellular activity. However, the mainstay of traditional RNA secondary structure prediction methods relies on thermos-dynamic models via free energy minimization, a laborious process that requires a lot of prior knowledge. Here, RNA secondary structure prediction using Wfold, an end-to-end deep learning-based approach, is suggested. Wfold is trained directly on annotated data and base-pairing criteria. It makes use of an image-like representation of RNA sequences, which an enhanced U-net incorporated with a transformer encoder can process effectively. Wfold eventually increases the accuracy of RNA secondary structure prediction by combining the benefits of self-attention mechanism's mining of long-range information with U-net's ability to gather local information. We compare Wfold's performance using RNA datasets that are within and across families. When trained and evaluated on different RNA families, it achieves a similar performance as the traditional methods, but dramatically outperforms the state-of-the-art methods on within-family datasets. Moreover, Wfold can also reliably forecast pseudoknots. The findings imply that Wfold may be useful for improving sequence alignment, functional annotations, and RNA structure modeling.

MAFNet: Multimodal Asymmetric Fusion Network for Radar Echo Extrapolation

Radar echo extrapolation (REE) is a crucial method for convective nowcasting, and current deep learning (DL)-based methods for REE have shown significant potential in severe weather forecasting tasks. Existing DL-based REE methods use extensive historical radar data to learn the evolution patterns of echoes, they tend to suffer from low accuracy. This is because data of radar modality face difficulty adequately representing the state of weather systems. Inspired by multimodal learning and traditional numerical weather prediction (NWP) methods, we propose a Multimodal Asymmetric Fusion Network (MAFNet) for REE, which uses data from radar modality to model echo evolution, and data from satellite and ground observation modalities to model the background field of weather systems, collectively guiding echo extrapolation. In the MAFNet, we first extract overall convective features through a global shared encoder (GSE), followed by two branches of local modality encoder (LME) and local correlation encoders (LCEs) that extract convective features from radar, satellite, and ground observation modalities. We employ an multimodal asymmetric fusion module (MAFM) to fuse multimodal features at different scales and feature levels, enhancing radar echo extrapolation performance. Additionally, to address the temporal resolution differences in multimodal data, we design a time alignment module based on dynamic time warping (DTW), which aligns multimodal feature sequences temporally. Experimental results demonstrate that compared to state-of-the-art (SOTA) models, the MAFNet achieves average improvements of 1.86% in CSI and 3.18% in HSS on the MeteoNet dataset, and average improvements of 4.84% in CSI and 2.38% in HSS on the RAIN-F dataset.

Prediction of regional water resources carrying capacity based on stochastic simulation: A case study of Beijing-Tianjin-Hebei Urban Agglomeration

Study regionBeijing-Tianjin-Hebei urban agglomeration in China Study focusThe prediction of water resources carrying capacity (WRCC) can provide an effective reference for the rational allocation and efficient utilization of water resources. Traditional prediction methods obtained a definite WRCC value but fail to reflect the uncertainty of WRCC changes and limit reference for the optimal allocation of water resources. To ensure the accuracy, availability and comprehensiveness of prediction, this paper adopts the improved principal component analysis (PCA) to screen indicators, and predicts the WRCC through the coupled model of Monte Carlo and Grey Wolf Optimization-Support Vector Machine(GWO-SVM), addressing single result issues and computational complexity. At the same time, various regulation schemes for sensitive indicators are designed to provide an effective guidance for the optimal allocation and sustainable use of water resources. New hydrological insights for the regionIn 2025, the probability of WRCC in Tianjin, Handan, Xingtai, Hengshui, Cangzhou, Langfang to maintain grade III is more than 80 %, and that in Beijing, Baoding, Tangshan, Qinhuangdao, Zhangjiakou, Chengde to reach grade IV is more than 50 %. The sensitivity analysis shows that the sensitive indicators mainly focus on water supply and consumption, water use efficiency and pollutant gas emissions. The WRCC can be further improved under different schemes. The results can provide effective guidance for the optimal allocation of water resources and maintain sustainable economic and social development in Beijing-Tianjin-Hebei urban agglomeration.

BioPathNet: Enhancing Link Prediction in Biomedical Knowledge Graphs through Path Representation Learning.

Understanding complex interactions in biomedical networks is crucial for advancements in biomedicine, but traditional link prediction (LP) methods are limited in capturing this complexity. Representation-based learning techniques improve prediction accuracy by mapping nodes to low-dimensional embeddings, yet they often struggle with interpretability and scalability. We present BioPathNet, a novel graph neural network framework based on the Neural Bellman-Ford Network (NBFNet), addressing these limitations through path-based reasoning for LP in biomedical knowledge graphs. Unlike node-embedding frameworks, BioPathNet learns representations between node pairs by considering all relations along paths, enhancing prediction accuracy and interpretability. This allows visualization of influential paths and facilitates biological validation. BioPathNet leverages a background regulatory graph (BRG) for enhanced message passing and uses stringent negative sampling to improve precision. In evaluations across various LP tasks, such as gene function annotation, drug-disease indication, synthetic lethality, and lncRNA-mRNA interaction prediction, BioPathNet consistently outperformed shallow node embedding methods, relational graph neural networks and task-specific state-of-the-art methods, demonstrating robust performance and versatility. Our study predicts novel drug indications for diseases like acute lymphoblastic leukemia (ALL) and Alzheimer's, validated by medical experts and clinical trials. We also identified new synthetic lethality gene pairs and regulatory interactions involving lncRNAs and target genes, confirmed through literature reviews. BioPathNet's interpretability will enable researchers to trace prediction paths and gain molecular insights, making it a valuable tool for drug discovery, personalized medicine and biology in general.

Remaining useful life prediction based on time-series features and conformalized quantile regression

Abstract The remaining useful life (RUL) prediction is a key task in the field of prognostics and health management (PHM) and plays a crucial role in preventive maintenance tasks. Traditional prediction methods have mostly focused on point prediction issues, neglecting the uncertain factors in the prediction task, thus failing to ensure the credibility of the prediction. In light of this, this paper focuses on improving the accuracy of point prediction models for RUL and interval prediction issues, proposing the introduction of multi-scale convolutional neural networks (MCNN), decomposed time-sequential linear layers (DL), and conformal quantile regression (CQR) techniques into the RUL prediction task of aero-engines. The aim is to provide timely and accurate failure warnings for aero engines, effectively ensure their reliability and safety, and reduce maintenance costs throughout their life cycle. In response to the limitations of current point prediction models in capturing the temporal features of life data, a MCNN-DL-based RUL prediction model is proposed to capture life data's long-term trends and local variations for precise point predictions. Furthermore, an interval estimation approach for RUL is presented, which integrates the MCNN-DL model with CQR to account for prediction uncertainty. Finally, the method in this paper is verified using the CMAPSS dataset, and the results show that the method has achieved excellent results in both RUL point prediction and interval prediction tasks.

Enhancing Flight Delay Predictions Using Network Centrality Measures

Accurately predicting flight delays remains a significant challenge in the aviation industry due to the complexity and interconnectivity of its operations. The traditional prediction methods often rely on meteorological conditions, such as temperature, humidity, and dew point, as well as flight-specific data like departure and arrival times. However, these predictors frequently fail to capture the nuanced dynamics that lead to delays. This paper introduces network centrality measures as novel predictors to enhance the binary classification of flight arrival delays. Additionally, it emphasizes the use of tree-based ensemble models, specifically random forest, gradient boosting, and CatBoost, which are recognized for their superior ability to model complex relationships compared to single classifiers. Empirical testing shows that incorporating centrality measures improves the models’ average performance, with random forest being the most effective, achieving an accuracy rate of 86.2%, surpassing the baseline by 1.7%.

Multivariate quality prediction of thin-walled parts machining using multi-task parallel deep transfer learning

Multivariate machining quality prediction of thin-walled parts with multiple machining features is a complex problem due to different data distribution between training and unlabelled test samples. Traditional quality prediction methods ignore the correlation of multiple quality labels and do not consider changes in data distribution, resulting in low accuracy of multivariate quality prediction. Therefore, a multivariate quality prediction method using multi-task parallel deep transfer learning is proposed to solve this problem. Specifically, a multi-output quality prediction model of cross-machining features is constructed through the joint design of multi-task parallel learning and deep transfer learning. Furthermore, a domain matcher is designed to form multiple transfer strategies, which can be used for dynamic matching of multi-source and multi-target machining features with multiple quality labels. The domain invariant data features through dynamic domain adaptation are extracted to deal with data distribution discrepancy between the source and target domains. Finally, the results of multiple comparison experiments show that the proposed method can effectively achieve the accurate quality prediction of the target domain with unlabelled labels and different distributions. Compared with the traditional methods, the proposed method has improved by 8.34%, 7.14%, and 9.09%, respectively, in MAE, RMSE, and score on average.

Productivity Prediction Model of Tight Oil Reservoir Based on Particle Swarm Optimization–Back Propagation Neural Network

Single-well productivity is a crucial metric for assessing the effectiveness of petroleum reservoir development. The accurate prediction of productivity is essential for achieving the efficient and economical development of oil–gas reservoirs. Traditional productivity prediction methods (empirical formulae and numerical simulation) are limited to specific reservoir types. There are few influencing factors, and a large number of ideal assumptions are made for the assumed conditions when predicting productivity. The application scenario is ideal. However, in tight oil reservoirs, numerous factors affect productivity, and their interactions exhibit significant complexity. Continuing to use traditional reservoir productivity prediction methods may result in significant calculation errors and lead to economic losses in oilfield development. To enhance the accuracy of tight reservoir productivity predictions and achieve economical and efficient development, this paper investigates the tight reservoir in the WZ block of the Beibuwan area, considering the impact of geological and engineering factors on productivity; the random forest tree and Spearman correlation coefficient are used to analyze the main influencing factors of productivity. The back propagation neural network optimized by particle swarm optimization was employed to develop a productivity prediction model (PSO-BP model) for offshore deep and ultra-deep tight reservoirs. The actual test well data of the oilfield are substituted into this model. The analysis results of the example application yielded an RMSE of 0.032, an MAE of 1.209, and an R2 value of 0.919. Compared with traditional productivity prediction methods, this study concludes that the model is both reasonable and practical. The calculation speed is faster and the calculation result is more accurate, which can greatly reduce productivity errors. The model constructed in this paper is well suited for predicting the productivity of tight reservoirs within the WZ block. It offers substantial guidance for predicting the productivity of similar reservoirs and supports the economical and efficient development of petroleum reservoirs.

Random projection ensemble conformal prediction for high-dimensional classification

In classification problems, many models with superior performance fail to provide confidence estimates or intervals for each prediction. This lack of reliability poses risks in real-world applications, making these models difficult to trust. Conformal prediction, as distribution-free and model-free approaches with finite-sample coverage guarantee, have recently been widely used to construct prediction sets for classification models. However, traditional conformal prediction methods only produce set-valued results without specifying a definitive predicted class. Particularly in complex settings, these methods fail to assist models in effectively addressing challenges such as high dimensionality, resulting in ambiguous prediction sets with low statistical efficiency, i.e. the prediction sets contain many false classes. In this study, a novel Ensemble Conformal Prediction algorithm based on Random Projection and a designed voting strategy, RPECP, is developed to tackle these challenges. Initially, a procedure for selecting the approximately oracle random projections and classifiers is executed to best leverage the internal information and structure of the data. Subsequently, based on the approximately oracle random projections and underlying classifiers, conformal prediction is performed on new test samples in a lower-dimensional space, resulting in multiple independent prediction sets. Finally, an accurate predicted class and a precise prediction set with high coverage and statistical efficiency are produced through a designed voting strategy. Compared to several base classifiers, RPECP obtain higher classification accuracy; against other conformal prediction algorithms, it achieves less ambiguous prediction sets with fewer false classes while guaranteeing high coverage. For illustration, this paper demonstrates RPECP's superiority over other methods in four cases: two high-dimensional settings and two real-world datasets.

Prediction of the internal corrosion rate for oil and gas pipelines and influence factor analysis with interpretable ensemble learning

Corrosion is one of the major threats to the safety and reliability of oil and gas pipelines, making accurate prediction of corrosion rate crucial for pipeline maintenance and repairment. Traditional prediction methods often ignore more critical factors and lack interpretability, which hinders the practical application. Here, an interpretable ensemble machine learning framework is proposed, not only improving prediction performance, but also enhancing interpretability for predicting the internal corrosion rate of oil and gas pipelines. In this work, ExtraTreeRegression model has demonstrated superior prediction accuracy relative to the other five machine learning models, and the determination coefficient of the ExtraTreeRegression model achieves 0.93 after feature engineering. Then, Shapley Additive exPlanations (SHAP) values is utilized to visually interpret the model locally and globally to help account for the contributions of the input features. Furthermore, the accumulated local effect (ALE) successfully explains how the features affect the internal corrosion rate of oil and gas pipelines. By collecting corrosion data of oil and gas pipeline and performing feature engineering and data preprocessing, we construct a comprehensive and reliable prediction model with interpretability. Experimental results demonstrate that the proposed interpretable ensemble machine learning approach outperforms other models in both accuracy and interpretability, providing valuable insights for pipeline management decisions.

Prediction of Human Microbe-Drug Association based on Layer Attention Graph Convolutional Network.

In this paper, we have proposed a novel model of layer attention graph convolutional network for microbe-drug association prediction. First, multiple biological data have been integrated into a heterogeneous network. Then, the heterogeneous network has been incorporated into a graph convolutional network to determine the embedded microbe and drug. Finally, the microbe-drug association scores have been obtained by decoding the embedding of microbe and drug based on the layer attention mechanism. To evaluate the performance of our proposed model, leave-one-out crossvalidation (LOOCV) and 5-fold cross-validation have been implemented on the two datasets of aBiofilm and MDAD. As a result, based on the aBiofilm dataset, our proposed model has attained areas under the curve (AUC) of 0.9178 and 0.9022 on global LOOCV and local LOOCV, respectively. Based on aBiofilm dataset, the proposed model has attained an AUC value of 0.9018 and 0.8902 on global LOOCV and local LOOCV, respectively. In addition, the average AUC and standard deviation of the proposed model for 5- fold cross-validation on the aBiofilm and MDAD datasets were 0.9141±6.8556e-04 and 0.8982±7.5868e-04, respectively. Also, two kinds of case studies have been further conducted to evaluate the proposed models. Traditional methods for microbe-drug association prediction are timeconsuming and laborious. Therefore, the computational model proposed was used to predict new microbe-drug associations. Several evaluation results have shown the proposed model to achieve satisfactory results and that it can play a role in drug development and precision medicine.

Analytical time-coordinated entry guidance for multi-hypersonic vehicles within three-dimensional corridor

A coordinated entry mission requires multiple hypersonic vehicles to arrive at the designated area simultaneously from different times and locations. To satisfy the coordinated entry mission, this paper investigates the coordinated entry guidance problem for multi-hypersonic vehicles. An analytical time-coordinated entry guidance method is proposed based on the algorithms of analytical prediction, profile correction, and coordination time determination. Firstly, considering the altitude variations of the trajectory, high-precision analytical solutions for the remaining flight time and range are derived based on quasi-equilibrium glide conditions. Compared to the traditional numerical integration prediction methods, this analytical prediction method achieves superior computational efficiency while ensuring high-precision prediction. Subsequently, a flight corridor based on longitudinal lift-to-drag ratio is constructed, extending from the traditional two-dimensional corridor to a three-dimensional one. Within the corridor, the analytical lift-to-drag ratio profile is corrected by minimizing the prediction errors, and the guidance commands for both angle of attack and bank angle can be generated directly without tracking law. With the above analytical solutions and algorithms, the coordinated entry guidance method enables rapid generation of guidance commands for each hypersonic vehicle. Finally, numerical simulations under dispersed launching time and launching location conditions are performed, and the simulation results verify the effectiveness and robustness of the proposed time-coordination entry guidance algorithm.

Implementing machine learning techniques for customer retention and churn prediction in telecommunications

This review paper explores the application of machine learning techniques in predicting customer churn and enhancing customer retention within the telecommunications industry. The paper begins by discussing the significance of customer churn, its causes, and the limitations of traditional churn prediction methods. It then delves into machine learning algorithms, including decision trees, support vector machines, and ensemble methods. It highlights their effectiveness in handling large and complex datasets typical of the telecom sector. The discussion extends to the challenges faced in data quality, model selection, implementation, and ethical considerations in using customer data for predictive analytics. The paper also compares machine learning models with traditional methods, emphasizing the advantages of scalability, accuracy, and real-time processing. Furthermore, it identifies potential innovations, such as improved data integration, interpretable models, and personalized retention strategies. Finally, the paper reflects on future trends, predicting the growing role of AI and machine learning in telecommunications, particularly in customer service automation and network optimization. The review underscores the importance of adopting machine learning to reduce churn and improve customer retention while considering the field's ethical implications and future opportunities. Keywords: Customer Churn Prediction, Machine Learning, Telecommunications, Customer Retention, Predictive Analytics, AI in Telecom

Droplet volume prediction methods in electrohydrodynamic jet printing based on multi-source data fusion

Electrohydrodynamic jet printing technology has garnered significant interest in additive manufacturing due to its advantages in higher resolution printing and greater viscosity material applicability. Nonetheless, the challenges associated with femtoliter-level droplets and the micrometer-scale printing space have rendered traditional optical observation methods for volume calculation impractical for integration into the printing systems. Obtaining the droplet volume accurately has become a formidable challenge in electrohydrodynamic printing. While existing research utilizes machine learning for the end-to-end prediction of process parameters to droplet volume, traditional prediction methods cannot achieve online prediction due to the complex spray states characteristic of electrohydrodynamic jet printing and pose integration difficulties within printing systems. This paper introduces a multi-source data fusion model suitable for electrohydrodynamic printing droplet volume prediction. The model employs the VGG network to extract features from Taylor cone images in the jetting state and synchronously fuse these with process parameter features. The fused features are then correlated with droplet volume labels through the MLP network for comprehensive model training. The performance of the proposed model has improved by 10% in prediction accuracy compared to single modal data. We integrated the prediction method into an electrohydrodynamic jet printing system and experimented with printing pixel-pit substrates. The results indicate that the prediction accuracy of the volume prediction system is over 92%. The printing efficiency has improved approximately 3 times compared to the traditional method, significantly enhancing overall performance.

The Utilization of Time Series Data Forecasting Techniques on Hospitals Pharmaceutical Inventory

Good inventory management is essential in the hospital industry to overcome inventory problems. Ineffective inventory prediction methods can lead to shortages or excess stock inventory. Ultimately, this can impact the budget and availability of pharmaceutical items in the hospital. Previous traditional prediction methods often show inaccuracies. This research utilizes ARIMA (Autoregressive Integrated Moving Average) and FB Prophet methods to predict the demand for pharmaceutical items in hospitals. In an attempt to evaluate the effectiveness of both methods, an experiment was conducted on five pharmaceutical items. The results showed that the ARIMA method produced better performance compared to the FB Prophet method, with the smallest error of 0.07310.

Recursive identification of kernel-based hammerstein model for online energy efficiency estimation of large-scale chemical plants

Accurate energy efficiency prediction is crucial for decision-making in large-scale chemical production, especially considering energy management and environmental concerns. The complexity of chemical processes, characterized by strong nonlinearities, dynamics, and uncertainties, challenges traditional prediction methods. This paper introduces KHSM (Kernel-based Hammerstein Subspace Model), which models nonlinear static and linear dynamic processes in series by embedding kernel function-based nonlinear mapping into subspace modeling for energy efficiency estimation. KHSM identifies nonlinear process models without explicit nonlinear mapping functions, constructing dynamic models that encompass both energy input and product output variables. An adaptive KHSM variant dynamically updates the energy efficiency model, improving accuracy by iteratively selecting and excluding samples. Comparative analysis shows KHSM outperforms existing methods, achieving a coefficient of determination (R2) closest to 1 and reducing the root-mean-square error (RMSE) by up to an order of magnitude, with the smallest mean and standard deviation. Additionally, modeling efficiency increases by an order of magnitude. The developed model enables ongoing energy efficiency estimation and proactive identification of efficiency trends. Validation through mathematical and real-world ethylene process cases demonstrates KHSM's efficacy and applicability. These estimates facilitate informed adjustments to energy management strategies and optimization of production approaches in energy-intensive chemical processes.

Utilizing the Kolmogorov-Arnold Networks for chiller energy consumption prediction in commercial building

Accurate prediction of chiller energy consumption is essential for optimizing energy usage and reducing operational costs in commercial buildings. Traditional predictive methods often struggle to capture the complex, nonlinear relationships inherent in energy consumption data. This study proposes the use of Kolmogorov-Arnold Networks (KAN) to address this challenge, leveraging their ability to model intricate nonlinear dynamics with high precision. The study introduces KAN as a novel application for real-world chiller energy prediction, using actual data obtained from a commercial building. The methodology involves comparing KAN's performance with Artificial Neural Networks (NN) and a hybrid metaheuristic algorithm combined with deep learning, namely the Teaching-Learning-Based Optimization with Deep Learning (TLBO-DL). The results show that KAN achieves an R2 value of 0.9465 and an RMSE of 6.1023, outperforming NN (R2: 0.9281, RMSE: 6.7709) and TLBO-DL (R2: 0.9366, RMSE: 6.2892). The novelty of this research lies in the innovative application of KAN to chiller energy consumption prediction, coupled with advanced parameter tuning and improved computational efficiency. This study not only demonstrates the superior accuracy of KAN but also contributes to the field by showcasing its practical utility and effectiveness in energy management systems.