Optimal Model Selection Research Articles

Comparative analysis of different model depths on convolutional neural network for handwritten digit recognition

In the rapidly advancing field of artificial intelligence, Convolutional Neural Networks (CNNs), as a representative method, have risen to prominence as a pivotal instrument for handling visual data. However, despite their widespread use, the impact of CNN depth on performance remains under-explored. This study delves into this aspect, evaluating the performance of CNN architectures with different depths - two-layer, four-layer, and five-layer - on the MNIST dataset, a version from the National Institute of Standards and Technology, a well-known benchmark dataset for handwritten digit recognition. Experimental results reveal that the four-layer model achieved the highest average accuracy of 99.76%, while the five-layer model, despite its additional complexity, only slightly trailed behind with a 99.73% accuracy rate. However, the five-layer model required a significantly longer training time. In conclusion, while deeper networks can increase accuracy, they can also introduce computational inefficiencies without significant gains in performance. This research provides a better understanding of CNN depth, guiding optimal model selection for image classification tasks.

An Adjustable Wireless Backhaul Link Selection Algorithm for LEO-UAV-Sensor-Based Internet of Remote Things Network.

Internet of Remote Things (IoRT) networks utilize the backhaul links between unmanned aerial vehicles (UAVs) and low-earth-orbit (LEO) satellites to transfer the massive data collected by sensors. However, the backhaul links change rapidly due to the fast movement of both the UAVs and the satellites, which is different from conventional wireless networks. Additionally, due to the various requirements of IoRT multiservices, the system performance should be comprehensively considered. Thus, an adjustable wireless backhaul link selection algorithm for a LEO-UAV-sensor-based IoRT network is proposed. Firstly, an optimization model for backhaul link selection is proposed. This model uses Q, which integrates the remaining service time and capacity as the objective function. Then, based on the snapshot method, the dynamic topology is converted into the static topology and a heuristic optimization algorithm is proposed to solve the backhaul link selection problem. Finally, the proposed algorithm is compared with two traditional algorithms, i.e., maximum service time and maximum capacity algorithms. Numerical simulation results show that the proposed model can achieve better system performance, and the overload of the satellites is more balanced. The algorithm can obtain a trade-off between remaining service time and capacity by dynamically adjusting model parameters. Thus, the adjustable backhaul link selection algorithm can apply to multiservice IoRT scenarios.

Spectrum of Complex Partial Performance Disorders in Patients with Mental Illnesses

Recent research highlights a high prevalence of minimal cerebral dysfunctions (MCD) in patients with mental disorders. Nonetheless, empirical understanding of minimal cerebral dysfunctions and diffuse neuropsychological impairments in adult patients remains limited. In our study, we examined 399 patients with diverse common mental disorders on minimal cerebral dysfunctions. 329 patients were in psychosomatic rehabilitation, 40 in forensic psychiatry, and 30 in an addiction ward of a psychiatric hospital at the time of the study. Symptom patterns and anamnesis of MCD and sociodemographic characteristics of the patients were recorded using structured questionnaires. Overall, about 29% of all patients reported MCD symptoms. Patients with MCD reported greater problem burden in everyday coping (attention, arithmetic, memory, sensitivity to noise, orientation) and social interaction (outsider, excitability) than patients without partial performance disorders. Patients with mental illness reported a heterogeneous pattern of minimal cerebral dysfunctions. These frequent disorders should be recognized in diagnostics and treatment. Further studies should investigate therapeutic approaches for MCD according to Baltes' Selective Optimization and Compensation model.

Analysis of Attack Detection on Log Access Servers Using Machine Learning Classification: Integrating Expert Labeling and Optimal Model Selection

Purpose: As the complexity and diversity of cyberattacks continue to grow, traditional security measures fall short in effectively countering these threats within web-based environments. Therefore, there is an urgent need to develop and implement innovative, advanced techniques tailored specifically to detect and address these evolving security risks within web applications.Methods: This research focuses on analyzing attack detection in log access servers using machine learning classification with two primary approaches: expert labeling integration and best model selection. Expert labeling determines whether log entries are safe or indicate an attack.Result: Validation in labeling was applied using different datasets to minimize errors and increase confidence in the resulting dataset. Experimental results show that the Decision Tree and Random Forest models have nearly identical accuracy rates, around 89.3%-89.4%, while the ANN model has an accuracy of 81%.Novelty: This study proposes a fusion of expert knowledge in labeling log entries with a rigorous process of selecting the best classification model. This integration has not been extensively explored in previous research, offering a novel approach to enhancing attack detection within web applications. The research contribution lies in the integration of expert security assessment and the selection of the best model for detecting attacks on server access logs, along with validating labels using various datasets from different log devices to enhance confidence in the analysis results.

Machine learning for power outage prediction during hurricanes: An extensive review

The surge of machine learning (ML) applications and increasing usage of data driven approach for resilience enhancement provide great opportunities for applying ML techniques to power outage prediction (POP) during hurricanes. Considering the substantial damage that wind hazards can cause to power system network (PSN) in countries such as the USA and China, researchers have devised numerous approaches for disaster prevention to reduce their susceptibility. Since the selection of optimal ML algorithms for POP during hurricanes is a complex task, this paper helps researchers easily comprehend the numerous prediction techniques used worldwide for power outages during hurricanes. The paper provides a brief review of several categories of hurricanes and the damage it causes to the PSN more specifically to distribution side, the necessary measures to improve the PSN resilience, and finally a sequential approach for optimal ML model selection techniques that researchers and engineers can use to improve the POP during hurricanes. In this study the effectiveness of ML algorithms has been explored with the help of performance evaluation metrics which serve as the basic deciding criteria for optimal model selection. This research also highlights some of the key issues and challenges of integrating machine learning algorithms with existing PSN.

Forest aboveground biomass estimation by GEDI and multi-source EO data fusion over Indian forest

ABSTRACT Monitoring changes in carbon stocks through forest biomass assessment is crucial for carbon cycle studies. However, challenges in obtaining timely and reliable ground measurements hinder creation of the spatially continuous maps of forest aboveground biomass density (AGBD). This study proposes an approach for generating spatially continuous maps of forest aboveground biomass density (AGBD) by combining Global Ecosystem Dynamics Investigation (GEDI) LiDAR-based data with open-access earth observation (EO) data. The key contribution of the study lies in the systematic evaluation of various model configurations to select the optimal model for AGBD generation. The evaluation considered various model configurations, including predictor sets, spatial resolution, beam selection, and sensitivity thresholds. We used a Random Forest model, trained through five-fold cross-validation on 80% of the total data, to estimate AGBD in the Indian forest region. Model performance was assessed using the 20% independent test dataset. Results, using Sentinel-1 and 2 predictors, yielded R2 values of 0.55 to 0.60 and RMSE of 48.5 to 56.3 Mg/ha. Incorporating forest and agroclimatic zone attributes improved performance (R2: 0.59 to 0.69, RMSE: 42.2 to 53.3 Mg/ha). The selection of the top 15 predictors, which favoured features from Sentinel-2, DEM, forest attributes, and agroclimatic zones, and GEDI data with sensitivity >0.98, yielded the optimal model with an R2 of 0.64 and RMSE of 46.59 Mg/ha. The results underscore the significance of incorporating attributes like forest and agro-climatic zones and the need for an optimal model selection considering predictor types and GEDI shot characteristics. The top-performing model is validated in Simdega, Jharkhand (R2: 0.74, RMSE: 39.3 Mg/ha), demonstrating the methodological potential of this approach. Overall, this study emphasizes the methodological prospects of integrating multi-source open-access EO data to produce spatially continuous aboveground biomass (AGB) maps through data fusion.

Optimal Scheme Selection Model of Distributed Photovoltaic Consumption in Smart City Based on Gaussian Mixture Model

In an effort to guarantee the stable action and photovoltaic absorption capability of the spread mesh after a macro digit of distributed solar panels are related to the mesh, this paper puts forward the definitive scheme selection model of distributed photovoltaic absorption in smart cities based on Gaussian mixture model. The principal constituent analysis arithmetic is used to process the chronicle data of distributed photovoltaic action, and the processed data are transmitted to the Gaussian compound model. According to this model, the short-term power situation of distributed photovoltaic is predicted, and the external ability and restrict requirements of the definitive programme of distributed photovoltaic consumption are determined according to the prediction fruits. The definitive programme of distributed photovoltaic consumption is selected by the approximate ideal solution sorting method. The measurement fruits show that the model has a good short-term power prediction effect of distributed photovoltaic, and the prediction fruits are in high agreement with the actual fruits, and the coverage rate and interval average bandwidth fruits are above 0.918 and below 0.11 respectively. The photovoltaic consumption of the definitive programme is above 130%.

Optimization Model for Selection of the Offshore Fleet Structure

Over the past century, with an accelerated increase in world energy demand, oil remained the most intriguing energy source, while oil exploration and production evolved from an onshore to an offshore exploration and production facility. In logistic and operational planning, the supply of offshore installations is the primary dependent of offshore supply vessels (OSVs). OSVs are classified into two main specialized groups, Anchor Handling Tug Supply (AHTS) and Platform Supply Vessels (PSVs). The establishment and maintenance of offshore oil installations require considerable financial resources for the hire of OSVs, which increases the oil exploration and production operator’s overall budget. Planning the offshore fleet structure aims to reduce expenses and increase the utilization of OSVs. This work will analyze the actual data of the southwestern African offshore market based on 24/7 working time and the use analysis of marine activities to define the parameters of the offshore fleet structure. The research results will be used to develop a methodology for modeling the optimal infrastructure of the offshore fleet. With statistical data of all support vessels’ activities and the use of linear programming (LP), it is possible to determine the vessel’s employment pattern and the optimal fleet structure by type and number of OSVs.

Abstract C068: Construction of preclinical study model for optimal anticancer drug selection using PDX model in pancreatic cancer

Abstract Background: Translational research using the patient-derived xenograft (PDX) model is progressing rapidly and is also becoming widespread in pancreatic cancer research. Since PDX models retain the characteristics and the heterogeneity of the primary tumor, PDX models can be reflected in the effectiveness of the anticancer drug in human. The purpose of this study was to establish the PDX library of pancreatic cancer for preclinical study. Methods: The tissues of the pancreatic cancer were obtained from patients who underwent surgery at our hospital. Two PDX models were established using immunodeficient rats. One case was effective in gemcitabine plus nab-paclitaxel (GnP) therapy, and the other was ineffective in GnP therapy in preoperative treatment. 5-FU, CPT-11, L-OHP, GEM, and CDDP were administered to these PDX models once a week for 3 weeks. The primary endpoint was tumor shrinkage. Results: GEM was effective in PDX model in a case in which GnP therapy was effective in the clinical setting, in contrast, GEM was ineffective in PDX model in a case in which GnP was ineffective in the clinical setting. This result was correlated with actual clinical practice. Conclusions: In this study, single anticancer drug tests were performed on PDX models using immunodeficient rats. The efficacy of anticancer drugs was correlated with the clinical setting, indicating that our PDX models might be useful as a preclinical study model. We are currently conducting multi anticancer drug studies. Citation Format: Ryota Tanaka, Kenjiro Kimura, Naoki Tani, Shimpei Eguchi, Ken Kageyama, Shuhei Kushiyama, Shigeaki Kurihara, Masahiko Kinoshita, Kohei Nishio, Hiroji Shinkawa, Go Ohira, Akira Yamamoto, Takeaki Ishizawa. Construction of preclinical study model for optimal anticancer drug selection using PDX model in pancreatic cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr C068.

Residual LSTM-based short duration forecasting of polarization current for effective assessment of transformers insulation

The empirical application of polarization and depolarization current (PDC) measurement of transformers facilitates the extraction of critical insulation-sensitive parameters. This technique, rooted in time-domain dielectric response analysis, forms the bedrock for parameterization and insulation modeling. However, the inherently time-consuming nature of polarization current measurements renders them susceptible to data corruption. This article explores deep-learning-based short-duration techniques for forecasting polarization current to address this limitation. By incorporating spatial shortcuts, the residual long short-term memory (LSTM) network facilitates the seamless propagation of spatial and temporal gradients. Furthermore, the relative forecasting assessment of the proposed residual LSTM model’s performance is made against traditional LSTM, attention LSTM, gated recurrent units (GRU), and convolutional neural network (CNN) models. Thus, optimal model selection strategies are evaluated based on their capability to capture extended dependencies and short-term information present in the data. In addition, the Monte Carlo dropout prediction is employed to estimate uncertainty in polarization current forecasts. The findings demonstrate that the proposed residual LSTM network model for polarization current forecasting yields the lowest error metrics and maintains prediction consistency over the testing duration. Thus, the proposed approach significantly reduces PDC measurement time, providing an effective means to develop proactive maintenance strategies for evaluating the insulation condition of transformers.

Site selection and capacity determination of charging stations considering the uncertainty of users’ dynamic charging demands

Aiming at the problems of high investment and low efficiency in the planning and construction of electric vehicle (EV) charging stations in cities, an optimization model for site selection and capacity determination of charging stations considering the uncertainty of users’ dynamic charging demands is proposed. Firstly, based on the travel chain theory and the Origin-Destination (OD) matrix, the travel characteristics of EVs are studied, and the spatial and temporal distribution prediction model of EV charging load is established through the dynamic Dijkstra algorithm combined with the Monte Carlo method. Secondly, a site selection model for the charging station is established which takes the minimum annualized cost of the charging station operator and the annualized economic loss of the EV users as the goal. At the same time, the weighted Voronoi diagram and Adaptive Simulated Annealing Particle Swarm Optimization algorithm (ASPSO) are adopted to determine the optimal number/site selection and service scope of charging stations. Finally, an uncertain scenario set is introduced into the capacity determination model to describe the uncertainty of the users’ dynamic charging demands, and the robust optimization theory is utilized to solve the capacity of the charging station. A case study is carried out for the EV charging station planning problem in some urban areas of a northern city, and the validity of the model is verified.

Biobjective Optimization Model for Store Selection for BOPS Service in Omni-Channel Retail Chain

Biobjective Optimization Model for Store Selection for BOPS Service in Omni-Channel Retail Chain

Optimal route selection model in freight transport with customer collection approach using genetic and fuzzy algorithms

Optimal route selection model in freight transport with customer collection approach using genetic and fuzzy algorithms

Sentinel-2 image based smallholder crops classification and accuracy assessment by UAV data

Timely and accurate extraction of crop planting units plays a critical role in crop yield estimation, soil management, food supplies, and disaster warnings. However, precisely mapping crop types is challenging in smallholder farming systems due to heterogeneous and mixed pixels, where field sizes are small, and crop types are very diverse. In this paper, the crop type system of Huocheng County in northwest China as an example area, crop classification feature variables are constructed using Sentinel-2 remote sensing images, and combining the ReliefF algorithm, three feature selection classification models are established by 3092 crop type sampling points data. The crop planting units are extracted using pixel-based and object-based classification methods, respectively, and unmanned aerial vehicle (UAV) data assess the accuracy as ground truth. The Sentinel-2 image based smallholder crops classification results indicate that: The effective combination of optimal input feature variables selection and classification models significantly improves crop classification accuracy. This is particularly evident when integrating this approach into pixel-based classification, addressing challenges such as low pixel classification accuracy in regions of this type. The validation results from UAV data also confirm these findings. At validation points 2, 3, and 4, pixel-based classification demonstrates significantly higher accuracy compared to object-based classification, with pixel-based classification accuracy improving by 9.16%, 7.83%, and 31.83%, respectively. Although the UAV validation accuracy does not reach a high level, pixel-based classification remains the optimal choice for smallholder crop classification. This method offers new insights and references for research related to the classification of smallholder crops with complex planting structures.

Application of singular value decomposition analysis: Insights into the complex mechanisms of amyloidogenesis

Amyloidogenesis, with its multifaceted nature spanning from peptide self-assembly to membrane-mediated structural transitions, presents a significant challenge for the interdisciplinary scientific community. Here, we emphasize on how Singular Value Decomposition (SVD) can be employed to reveal hidden patterns and dominant modes of interaction that govern the complex process of amyloidogenesis. We first utilize SVD analysis on Circular Dichroism (CD) spectral datasets to identify the intermediate structural species emerging during peptide-membrane interactions and to determine binding constants more precisely than conventional methods. We investigate the monomer loss kinetics associated with peptide self-assembly using Nuclear Magnetic Resonance (NMR) dataset and determine the global kinetic parameters through SVD. Furthermore, we explore the seeded growth of amyloid fibrils by analyzing a time-dependent NMR dataset, shedding light on the kinetic intricacies of this process. Our analysis uncovers two distinct states in the aggregation of Aβ40 and pinpoints key residues responsible for this seeded growth. To strengthen our findings and enhance their robustness, we validate those using simulated data, thereby highlighting the physical interpretations derived from SVD. Overall, SVD analysis offers a model-free, global kinetic perspective, enabling the selection of optimal kinetic models. This study not only contributes valuable insights into the dynamics but also highlights the versatility of SVD in unravelling complex processes of amyloidogenesis.

A Network-based Optimization Model for Effective Target Selection

A Network-based Optimization Model for Effective Target Selection

Data-Informed Discovery of High-Performance Cu-Ligand Catalysts for Acetylene Hydrochlorination

Understanding the relationship between catalyst structure and performance is essential for the rational design of active sites for high-performance Cu-ligand catalysts for acetylene hydrochlorination. Herein, two descriptors are informed by data to screen efficient ligands and predict high-performance Cu-ligand catalysts: (1) the reduction potential of Cu2+-Cu+ and catalytic performance were correlated in a volcano-like manner; (2) the C≡C bond length of C2H2 and catalytic performance were linearly related within the Cu2+-Cu+ reduction limit. By using the two descriptors, the MPPO (3-methyl-1-phenyl-2-phospholene 1-oxide) ligand was successfully identified and developed into an effective Cu-MPPO catalyst. The conversion of C2H2 of 96.11 % and the selectivity of VCM of 99 % were obtained under the conditions of 180 h−1, 180 °C, and VHCl: VC2H2 = 1.2. The C2H2 conversion was still above 90 % after running for more than 600 h at 50 h−1. The excellent activity and exceptional stability of the Cu-MPPO catalyst validate the developed model for optimal ligand selection. Detailed structural characterization and density function theory calculations of the designed catalysts were performed to reveal the relationship between catalyst structure and activity. The results indicated that the addition of MPPO altered the local microenvironment of Cu. The simultaneous increased electron density around Cu and Cl resulted in strengthened HCl adsorption and weakened acetylene adsorption, inhibiting the generation of carbon deposition. In addition, MPPO contributed to suppressing the reduction of Cu species, thus stabilizing the active species. This research provides valuable insights and reference strategies for screening efficient ligands in Cu-ligand catalysts.

Dynamic joint sensor selection and maintenance optimization in partially observable deteriorating systems

We consider a degrading system with costly but silent failures. The system can be partially observed using a set of heterogeneous noisy sensors at a given cost, where each sensor noise level is a function of the system state. The goal is to devise a dynamic joint sensor selection and maintenance optimization model to minimize the total expected discounted cost of the system. We develop an infinite-horizon discrete-time partially observable Markov decision process model to dynamically prescribe the sensor subsets for system monitoring and the timing of maintenance activities. We perform numerical experiments for a system with up to ten sensors and compare the proposed optimal policy with four heuristic policies. We observe that the proposed approach consistently outperforms all heuristic policies considered. Through a comprehensive analysis of numerical experiments, our study shows the efficacy of the proposed model’s optimal policy in strategically selecting sensors to achieve precise system state estimation while managing costs. Through these experiments, we find that it is optimal to use lower-cost, higher-variance sensors and often fuse them for more accurate estimations. However, the choice depends on the system state and sensor quality distribution. Importantly, our model excels in determining the best action by considering these factors, underscoring its practical applicability.

Power load combination forecasting system based on longitudinal data selection

In the current situation of energy supply shortage and surging demand, effective and stable load forecasting is essential to ensure reliable power supply and the security of the power system. However, due to some factors such as periodicity and seasonality, the power load sequence shows complex nonlinear characteristics. Meanwhile, the current load forecasting lacks the ability to explore the data deeply, and it is difficult to accurately predict the short-term trend and fluctuation range. To remedy these limitations, this study proposes a hybrid point-interval prediction system (HPILS). The system integrates data preprocessing, optimal model selection, multi-objective optimization combination and interval prediction modules. To verify the performance of the proposed system, four load data sets in Australia are used as examples to conduct experiments. The experimental results demonstrate that HPILS can effectively provide the predicting power load trend changes and fluctuation ranges. Specifically, compared with the benchmark model, HPILS has a 13.47% ∼ 67.89% improvement in point prediction and a 1.67% ∼ 72.08% improvement in interval prediction. In addition, a series of discussion tests are performed to verify the superiority of the proposed system and further confirm the validity of our proposed system.

A comparative analysis: optimal node selection in large data block transmission in VANET using various node relay optimization algorithms

VANETs (Vehicular Ad-hoc Networks) are a kind of Mobile Ad-hoc Networks (MANETs) with increased mobility as well as occasional geographic changes. Clustering is used in VANETs to divide networks into groups of adaptable vehicles to optimize routing, information gathering, as well as routing. As large data block transmission contains several stages and inside those, one of the important stages such as optimal model selection was these selections of quintessential features helps in a further stage called data transmission much more effective. This stage will also bring energy optimization and also improves the effectiveness of the optimal path for data transmission. This paper brings a comparative analysis of this optimal relay node selection. Here, we use Linear Scaling based Red Colobuses Monkey technique and are contrasted with other state-of-the-art models over various measures. Analysis proves that the proposed optimal relay node selection technique boosts the data transmission model.