Incremental Model Research Articles

WSBCV: A data-driven cross-version defect model via multi-objective optimization and incremental representation learning

Cross-version defect prediction (CVDP) refers to training an excellent model on tagged defect data in a previously released project version, and then performing defect prediction on an unlabeled instance module in the current version. Nevertheless, the complicated internal intrinsic construction hidden behind the code defects makes it difficult for the previous cross-version defect models to capture more discriminative software features, and seriously restrains the CVDP performance. In this study, we propose an intelligent data-driven CVDP model named WSBCV based on multi-objective optimization and incremental representation learning. We firstly leverage an advanced deep generation adversarial network – WGAN-GP (Wasserstein GAN with Gradient Penalty) to perform data augmentation, including balancing defect classes and synthesizing abundant training instances. Secondly, a multi-objective SPEA/R (Strength Pareto-based Evolutionary Algorithm / Reference) feature selection optimization method is built to effectively search the fewest representative feature subsets while achieving the minimum error. Finally, a powerful defect predictor for CVDP based on the BLS (Broad Learning System) with incremental learning is built to learn excellent feature representations and achieve incremental online model update quickly. Experimental results across 32 cross-version pairs from 45 version demonstrate that the proposed SPEA/R, BLS and WSBCV all have statistically significant difference advantages compared to ten multi-objective feature selection approaches, six defect predictors and two CVDP models, respectively.

Improved stochastic configuration networks with vision patch fusion method for industrial image classification

This paper contributes to the advancement of stochastic configuration neural networks (SCN) in the field of visual applications. The proposed image classification randomized algorithm is an extension of deep stochastic configuration networks (DeepSCN), known as an improved SCN with vision patch fusion (Vi-SCN). Compared to existing two dimensional stochastic configuration networks (2DSCN) and DeepSCN, we have developed an incremental modeling method that employs a stochastic configuration patch fusion method. This method extracts randomly fused image features from three-channel high-resolution image data, improving the network's ability to extract features. Moreover, we have introduced a strategy for dynamically determining the depth structure of the network, enabling flexible adjustments to the network structure and the number of nodes in each layer based on the complexity of image recognition tasks. Through a series of comparisons on four image classification benchmark datasets, we assess the superior learning performance of our design compared to 2DSCN and DeepSCN. Furthermore, to evaluate the performance of Vi-SCN in practical visual application scenarios, we collect three industrial datasets with distinct characteristics. Vi-SCN demonstrate outstanding performance in all three tasks, showing its significant potential in image recognition.

The American ‘doc fix’: Incremental policy change and the growth of US healthcare spending

The extraordinary growth of healthcare spending in the United States necessitates an incremental model of institutional change. Efforts to reduce social spending in the 1980s introduced strict cost controls on healthcare, yet this regulation was annually abated with temporary legislation known as ‘doc fixes’. We show that the institutions set up to control spending were slowly undermined through three mechanisms: the transformation of wider political and economic problems into technical ones, the ‘layering’ effect of repeated budgetary appropriations, and the moral arguments that sustained incremental spending growth. This case illustrates how creative responses to budget constraints may produce incremental change, driven by ideas about how markets and welfare should work.

Hawk: Rapid Android Malware Detection Through Heterogeneous Graph Attention Networks.

Android is undergoing unprecedented malicious threats daily, but the existing methods for malware detection often fail to cope with evolving camouflage in malware. To address this issue, we present Hawk, a new malware detection framework for evolutionary Android applications. We model Android entities and behavioral relationships as a heterogeneous information network (HIN), exploiting its rich semantic meta-structures for specifying implicit higher order relationships. An incremental learning model is created to handle the applications that manifest dynamically, without the need for reconstructing the whole HIN and the subsequent embedding model. The model can pinpoint rapidly the proximity between a new application and existing in-sample applications and aggregate their numerical embeddings under various semantics. Our experiments examine more than 80 860 malicious and 100 375 benign applications developed over a period of seven years, showing that Hawk achieves the highest detection accuracy against baselines and takes only 3.5 ms on average to detect an out-of-sample application, with the accelerated training time of 50× faster than the existing approach.

Complex Continuous Action Iterated Dilemma With Incremental Dynamic Model

Complex Continuous Action Iterated Dilemma With Incremental Dynamic Model

Experimental Investigation of an Incremental Contact Model for Hyperelastic Solids Using an In Situ Optical Interferometric Technique

The intricacies of rough surface contact are amplified by hyperelastic materials, primarily due to nonlinear enhancement caused by stress concentration. In previous studies, we proposed an incremental contact model for hyperelastic materials based on the tangent modulus and validated it through finite element simulations. This study proceeds with the experimental validation of the model. Initially, four hyperelastic rough surfaces were scanned and stitched together using a white light interferometer to obtain the whole surface topography. Subsequently, in situ optical interferometric techniques precisely measured the actual contact areas between these four samples and quartz glass, establishing the relationship between the load and contact area. Finally, by incorporating the surface topography into the incremental contact model for hyperelastic materials using profile theory, predictions of the relationship between load and contact area were made and compared with the experimental results. Significant agreement was found within nearly 90% of the relative contact area, which validated the model’s efficacy. The importance of this model extends to practical domains, such as wear, sealing, and contact surface safety research.

HIPPO: HIstogram-based Pseudo-POtential for scoring protein-ssRNA fragment-based docking poses

BackgroundThe RNA-Recognition motif (RRM) is a protein domain that binds single-stranded RNA (ssRNA) and is present in as much as 2% of the human genome. Despite this important role in biology, RRM-ssRNA interactions are very challenging to study on the structural level because of the remarkable flexibility of ssRNA. In the absence of atomic-level experimental data, the only method able to predict the 3D structure of protein-ssRNA complexes with any degree of accuracy is ssRNA’TTRACT, an ssRNA fragment-based docking approach using ATTRACT. However, since ATTRACT parameters are not ssRNA-specific and were determined in 2010, there is substantial opportunity for enhancement.ResultsHere we present HIPPO, a composite RRM-ssRNA scoring potential derived analytically from contact frequencies in near-native versus non-native docking models. HIPPO consists of a consensus of four distinct potentials, each extracted from a distinct reference pool of protein-trinucleotide docking decoys. To score a docking pose with one potential, for each pair of RNA–protein coarse-grained bead types, each contact is awarded or penalised according to the relative frequencies of this contact distance range among the correct and incorrect poses of the reference pool. Validated on a fragment-based docking benchmark of 57 experimentally solved RRM-ssRNA complexes, HIPPO achieved a threefold or higher enrichment for half of the fragments, versus only a quarter with the ATTRACT scoring function. In particular, HIPPO drastically improved the chance of very high enrichment (12-fold or higher), a scenario where the incremental modelling of entire ssRNA chains from fragments becomes viable. However, for the latter result, more research is needed to make it directly practically applicable. Regardless, our approach already improves upon the state of the art in RRM-ssRNA modelling and is in principle extendable to other types of protein-nucleic acid interactions.

Influential Exemplar Replay for Incremental Learning in Recommender Systems

Personalized recommender systems have found widespread applications for effective information filtering. Conventional models engage in knowledge mining within the static setting to reconstruct singular historical data. Nonetheless, the dynamics of real-world environments are in a constant state of flux, rendering acquired model knowledge inadequate for accommodating emergent trends and thus leading to notable recommendation performance decline. Given the typically prohibitive cost of exhaustive model retraining, it has emerged to study incremental learning for recommender systems with ever-growing data. In this paper, we propose an effective model-agnostic framework, namely INFluential Exemplar Replay (INFER). INFER facilitates recommender models in retaining the earlier assimilated knowledge, e.g., users' enduring preferences, while concurrently accommodating evolving trends manifested in users' new interaction behaviors. We commence with a vanilla implementation that centers on identifying the most representative data samples for effective consolidation of early knowledge. Subsequently, we propose an advanced solution, namely INFERONCE, to optimize the computational overhead associated with the vanilla implementation. Extensive experiments on four prototypical backbone models, two classic recommendation tasks, and four widely used benchmarks consistently demonstrate the effectiveness of our method as well as its compatibility for extending to several incremental recommender models.

Pre-trained Online Contrastive Learning for Insurance Fraud Detection

Medical insurance fraud has always been a crucial challenge in the field of healthcare industry. Existing fraud detection models mostly focus on offline learning scenes. However, fraud patterns are constantly evolving, making it difficult for models trained on past data to detect newly emerging fraud patterns, posing a severe challenge in medical fraud detection. Moreover, current incremental learning models are mostly designed to address catastrophic forgetting, but often exhibit suboptimal performance in fraud detection. To address this challenge, this paper proposes an innovative online learning method for medical insurance fraud detection, named POCL. This method combines contrastive learning pre-training with online updating strategies. In the pre-training stage, we leverage contrastive learning pre-training to learn on historical data, enabling deep feature learning and obtaining rich risk representations. In the online learning stage, we adopt a Temporal Memory Aware Synapses online updating strategy, allowing the model to perform incremental learning and optimization based on continuously emerging new data. This ensures timely adaptation to fraud patterns and reduces forgetting of past knowledge. Our model undergoes extensive experiments and evaluations on real-world insurance fraud datasets. The results demonstrate our model has significant advantages in accuracy compared to the state-of-the-art baseline methods, while also exhibiting lower running time and space consumption. Our sources are released at https://github.com/finint/POCL.

Time-delay-aware power coordinated control approach for series hybrid electric vehicles

Series hybrid electric vehicles (S-HEVs) have attracted extensive attention thanks to the simple structure and excellent efficiency, which power transfer rely on high-voltage microgrids consisting of engine-generator set, battery, motor, and other actuators. However, with possible communication time delays (CTDs), the interaction between the actuators and control unit in terms of state and control command is desynchronize, resulting in the unstable operation of S-HEVs, such as engine stalls and bus voltage jitter. This presents a great challenge to the power coordinated control approach (PCCA) design that controls different actuators. Motivated by the mentioned above, this paper presents a time-delay-aware PCCA to achieve the stable driving of S-HEVs. First, an integrated powertrain control-oriented model that incorporates communication is built. Second, an increment model predictive control method (MPC) is designed. And then by leveraging the Lyapunov-Razumikhin stability theorem, a stability condition is derived by defining a parameterization weight matrix and known upper bound of CTDs, reducing the complexity of parameter designing. Additionally, an adaptive mechanism is designed to extend application of the proposed approach in unknown CTDs condition. Finally, the hardware-in-the-loop experiment is performed, and the results demonstrate that the presented approach in the mean and root mean square of the engine speed error and DC-bus voltage error decrease 39.87%, 43.85%, 44.67%, and 79.94% than nominal MPC in one period CTDs condition, respectively. And for the larger CTDs condition, the proposed method also shows efficient robustness. These substantiate that the proposed control method effectively improves vehicle running stability when encountering CTDs.

Incremental cost analysis model of distribution network based on economic dispatch of distributed new-energy storage system

Incremental cost analysis model of distribution network based on economic dispatch of distributed new-energy storage system

Topographic Elevation’s Impact on Local Climate and Extreme Rainfall: A Case Study of Zhengzhou, Henan

The topography significantly influences local climate precipitation and the intensity of precipitation events, yet the specific differences in its elevational effects require further understanding. This study focuses on precipitation in Zhengzhou City, Henan Province, utilizing hourly data and a topographic elevation precipitation increment model to assess the impact of topography on local climate precipitation and extreme heavy rainfall events. The results indicate that the daily precipitation attributed to topographic elevation in Zhengzhou in July was 0.21 mm, accounting for 4.9% of the total precipitation. In the extreme heavy rainfall event on 20 July 2021 (“7.20” event), the precipitation due to topographic elevation reaches 48.7 mm, constituting 15.8% of the total precipitation. Additionally, numerical simulations using the Weather Research and Forecasting (WRF) model for the 20–21 July 2021 rainfall event in Zhengzhou show that the WRF model effectively reproduces the spatiotemporal characteristics of the precipitation process. The simulated topographic elevation precipitation intensity is 49.8 mm/day, accounting for 16.6% of daily precipitation, closely resembling observational data. Sensitivity experiments further reveal that reducing the heights of the Taihang Mountains and Funiu Mountains weakens the low-level easterly winds around Zhengzhou. Consequently, as the center of the heavy rainfall shifts northward or westward, the intensity of topographic elevation-induced precipitation decreases to 7.3 mm/day and 12.9 mm/day.

Model Predictive Control of Counter-Rotating Motors for Underwater Vehicles Considering Unbalanced Load Variation

The propulsion system for underwater vehicles, driven by a counter-rotating permanent magnet synchronous motor (CRPMSM), can enhance the operational stability and efficiency of the vehicle. Due to the influence of complex underwater flows, the load imbalance of CRPMSM’s dual counter-rotating rotors may lead to severe issues of dual-rotor desynchronization rotation. Combining traditional vector control (VC) with master-slave control strategies can address the desynchronization problem when CRPMSM’s load changes. However, it results in significant speed fluctuations and a long transition time during the transition from load disturbance to synchronous rotation. This paper introduces a model predictive control (MPC) strategy to effectively resolve this issue. The incremental MPC model is established based on the mathematical model of CRPMSM in the dq coordinate system. The predictive control system forecasts the d- and q-axis components of stator currents for the next four control cycles. It selects the optimal control increments to minimize the cost function based on current predictions and different inverter voltage states. The obtained optimal d- and q-axis components of stator voltage are used to control CRPMSM under unbalanced load disturbances. Simulation results demonstrate that, compared to the VC strategy, CRPMSM utilizing the MPC strategy exhibits better dynamic performance with faster speed response and reduced torque fluctuations during load and speed variations.

Fine tuned personalized machine learning models to detect insomnia risk based on data from a smart bed platform.

Insomnia causes serious adverse health effects and is estimated to affect 10-30% of the worldwide population. This study leverages personalized fine-tuned machine learning algorithms to detect insomnia risk based on questionnaire and longitudinal objective sleep data collected by a smart bed platform. Users of the Sleep Number smart bed were invited to participate in an IRB approved study which required them to respond to four questionnaires (which included the Insomnia Severity Index; ISI) administered 6 weeks apart from each other in the period from November 2021 to March 2022. For 1,489 participants who completed at least 3 questionnaires, objective data (which includes sleep/wake and cardio-respiratory metrics) collected by the platform were queried for analysis. An incremental, passive-aggressive machine learning model was used to detect insomnia risk which was defined by the ISI exceeding a given threshold. Three ISI thresholds (8, 10, and 15) were considered. The incremental model is advantageous because it allows personalized fine-tuning by adding individual training data to a generic model. The generic model, without personalizing, resulted in an area under the receiving-operating curve (AUC) of about 0.5 for each ISI threshold. The personalized fine-tuning with the data of just five sleep sessions from the individual for whom the model is being personalized resulted in AUCs exceeding 0.8 for all ISI thresholds. Interestingly, no further AUC enhancements resulted by adding personalized data exceeding ten sessions. These are encouraging results motivating further investigation into the application of personalized fine tuning machine learning to detect insomnia risk based on longitudinal sleep data and the extension of this paradigm to sleep medicine.

Significant enhancement of Jc and flux pinning mechanism of MgB2 doped with PEG-polymers

Compared with traditional low-temperature superconductors, the advantage of MgB2 lies in its critical temperature close to 40 K, which provides potential application at 20 K. In order to further improve its superconducting current carrying capacity at 20 K, a series of PEG polymer (including PEG2000, PEG4000, and PEG6000) doped MgB2 with doping level of 15 wt% were prepared by combining hot-pressing sintering with chemical solution route. The best result is achieved in PEG4000 doped sample, with a Jc at 20 K and 2 T reaching 51,000 A/cm2, increased by 42.8 % in comparison with the undoped sample. Study on the mechanism of the improvement reveals that organic carbon doping enhances the irreversibility field and consequently the flux pinning force. In response to the deviation of MgB2 pinning characteristics from Kramer theory, an irreversibility field increment (IFI) model was proposed which provides a satisfactory explanation for the experimental results. Our results demonstrate that the combination of hot-pressing sintering and organic carbon doping is an effective means to enhance the superconducting performance of MgB2 at 20 K for practical applications.

Towards Effective Solid Waste Management: A Mobile Application for Coordinated Waste Collection and User-official Interaction

Solid Waste Management is an especially important task related to human health and the environment. Due to ineffective scheduled date & time, poor communication between waste collecting institutions and local house owners, people are compelled to throw waste on streets which is not good. Even if there is a routine, people tend to miss the schedule. Our aim is to develop an application for mobile phones, which consists of two parties- the user and waste management officials, where the second one acts as reminders. Officials will send a notification to the user, signaling that they are at a certain checkpoint near the user and the user can now throw waste properly and not on the streets. An incremental model was used throughout our project; basic requirements are fulfilled first and then iterated to create the final product. The proposed application includes two portals for whether you are user or waste management personnel. This application helps to improve the coordination between clients and collectors and determines whether the waste in an area has been collected or not. The survey conducted in this study involved consulting the Environment and Agricultural Department of Kathmandu Metropolitan City, which highlighted the significance of a notifying application. This application addresses the issue of uncoordinated waste disposal by providing users with information about collection schedules, leading to better waste management practices and reduced unsystematic garbage disposal.

Pollution Characteristics and Risk Assessment of Polycyclic Aromatic Hydrocarbons in Soils of Guangzhou

In order to comprehensively study the pollution characteristics of polycyclic aromatic hydrocarbons （PAHs） in soils of Guangzhou， 222 topsoil samples were collected and analyzed. The ecological risk of soil PAHs pollution was evaluated using the effect interval low/median method （ERL/ERM） and the （BaP） toxicity equivalent method， and the health risk of soil PAHs pollution was evaluated using the lifelong cancer risk increment model. The source of PAHs was analyzed using the characteristic compound ratio method and PMF model. The results indicated that： the content of surface soil （∑16PAHs） in Guangzhou was 38-11 115 μg·kg-1， with an average of 526 μg·kg-1， and 16 types of polycyclic aromatic hydrocarbon monomers showed strong variation. There was a certain degree of ecological risk of PAHs in Guangzhou， and there was already a significant ecological risk of PAHs pollution in individual sampling points， which were generally in a state of mild pollution. Based on the results of the health risk assessment， the contribution rates of total cancer risk in both adults and children were presented as follows： skin contact > ingestion of soil > respiratory intake. The health risk of children was greater than that of adults， and the overall health risk was within an acceptable range. Source analysis showed that the main sources of soil PAHs in Guangzhou were coal （37.1%）； diesel （32%）； coking （17.3%）； and mixed sources of traffic emissions， biomass combustion， and petrochemical product volatilization （13.6%）. The overall source of soil PAHs belonged to mixed sources. The research results have enriched our understanding of the pollution status of PAHs in the surface soil of Guangzhou and are helpful in promoting soil pollution prevention and control actions.

Guidelines to use the ICSM for developing quantum-classical systems

Quantum computing is the turning point that represents a revolution in software development that will make it possible to solve those problems unsolvable with classical computing. Just as in other milestones in the history of software development, such as the adoption of object-oriented systems, where new software development processes and new life cycles emerged, with the quantum computing revolution, a new life cycle for quantum and hybrid software systems is needed. Although there are some life cycle proposals for quantum software systems, most of them do not comprehensively address the specific needs of these systems. In this paper, a quantum life cycle proposal is presented adapted from the Incremental Commitment Spiral Model (ICSM) and an example of its use is presented. nd that projection cannot exceed $\lfloor\frac{n}{2}\rfloor$.

Guidelines to use the ICSM for developing quantum-classical systems

Quantum computing is the turning point that represents a revolution in software development that will make it possible to solve those problems unsolvable with classical computing. Just as in other milestones in the history of software development, such as the adoption of object-oriented systems, where new software development processes and new life cycles emerged, with the quantum computing revolution, a new life cycle for quantum and hybrid software systems is needed. Although there are some life cycle proposals for quantum software systems, most of them do not comprehensively address the specific needs of these systems. In this paper, a quantum life cycle proposal is presented adapted from the Incremental Commitment Spiral Model (ICSM) and an example of its use is presented. nd that projection cannot exceed $\lfloor\frac{n}{2}\rfloor$.

Task incremental learning-driven Digital-Twin predictive modeling for customized metal forming product manufacturing process

Task incremental learning-driven Digital-Twin predictive modeling for customized metal forming product manufacturing process