Model Building Phase Research Articles

Multivariate SPC via sequential multiblock-PLS

The sequential multi-block partial least squares (SMB-PLS) is proposed for implementing a multivariate statistical process control scheme. This is of interest when the system is composed of several blocks following a sequential order and presenting correlated information, for instance, a raw material properties block followed by a process variables block that is manipulated according to raw material properties. The SMB-PLS uses orthogonalization to separate correlated information between blocks from orthogonal variations. This allows monitoring the system in different stages considering only the remaining orthogonal part in each block. Thus, the SMB-PLS increases the interpretability and process understanding in the model building (Phase I), since it provides a deep insight about the nature of the system variations. Besides, it prevents any special cause from propagating to subsequent blocks enabling their use in the model exploitation (Phase II). The methodology is applied to a real case study from a food manufacturing process.

Climate covariate selection influences MaxEnt model predictions and predictive accuracy under current and future climates

The performance and transferability of species distribution models (SDMs) depends on a number of ecological, biological, and methodological factors. There is a growing body of literature that explores how the choice of climate covariate combinations and model parameters can affect predictive performance, but relatively few that delve into covariate reduction methods and the optimisation of model parameters, and the resulting spatial and temporal transferability of those models. The present work used the citrus pest, Diaphorina citri Kuwayama (Hemiptera: Psyllidae), to illustrate how MaxEnt models trained on the insect’s native range in Asia varied in their predictions of climatic suitability across the introduced range when eight different covariate reduction methods were applied during model building. Additionally, it showed how model sensitivity varied across these different covariate combinations using three sets of independently validated occurrence points in the invaded range of the psyllid. Climatically suitable areas for D. citri differed by as much as two-fold between the best and worst-performing models in selected areas. Great care should be taken in the selection of the highest-performing predictor combinations and model parameter settings for SDMs, particularly in the case of invasive species where the assumption of environmental equilibrium is likely violated in the introduced range. Understanding how the predictive ability of SDMs can be influenced by the methodological choices made during the model building phase is vital to ensuring that ecological and invasion management programmes do not over- or underestimate climatic suitability and subsequent invasion risk.

Design of a news recommendation model based on collaborative filtering and LSTM

This paper aims to design a news recommendation model based on user preferences to address the issues in recommendation systems under large datasets. Initially, four datasetsclick_history, news, news_embedding, and user_predictwere integrated into a single table, followed by data cleaning and feature engineering. Due to the large volume of data, this paper proposes necessary data filtering for the training and testing sets, utilizing temporal data to construct user feature vectors and news feature vectors. One challenge is how to effectively integrate user preferences and news features into the model to avoid overfitting or underfitting. In the model design and building phase, different methods were attempted to merge the information of users and news. Ultimately, the user preference features were processed using a fully connected layer, and the news embedding vectors were handled using an LSTM model. These two data parts were then combined into another fully connected layer, using ReLU as the activation function and CELoss as the loss function. Subsequently, the model's hyperparameters were adjusted and evaluated, achieving favorable model performance. The prediction accuracy for recommending news to users in user_predict was calculated as an evaluation criterion. Finally, this paper proposes directions for generalization and optimization in three aspects: data processing, model design, and experimental design. This includes data processing methods, potential improvements or mechanisms that could be incorporated into the model, and hyperparameter tuning. The paper primarily proposes data filtering to solve the problem of excessive data scale, which may aid in addressing recommendation system issues under large-scale datasets.

Personalized Image Aesthetics Assessment based on Graph Neural Network and Collaborative Filtering

Personalized image aesthetics assessment aims to capture individual aesthetic preferences, which are influenced by image aesthetic attributes and user demographic attributes. The interaction of attributes facilitates the determination of users' aesthetic preferences for images. Therefore, we define two forms of attribute interactions: external-interactions and internal-interactions. The interaction of these two types of attributes is not considered in existing models. To address this drawback, we suggest a personalized image aesthetics assessment method based on graph neural network and collaborative filtering, which models and aggregates two types of attribute interactions in the graph structure for predicting personalized image aesthetics scores. Firstly, we designed an image aesthetic feature extraction phase for obtaining aesthetic attributes and distributions based on the aesthetic assessment of mass images. Secondly, we propose an aesthetic prior model-building phase with two basic processes: learning the aesthetic features of images and users' aesthetic viewpoints; learning users' preferences for images. This phase is accomplished through internal-interactions (using the graph's message passing mechanism) and external-interactions (using collaborative filtering). Finally, we fuse the post-interaction features and image aesthetic distribution features for personalized image aesthetic assessment. The performance of our designed method is outperformed by the state-of-the-art method, as seen from the experimental results. Furthermore, further studies verify the accuracy and validity of our model in providing improved prediction of users' aesthetic preferences.

Multi-Strategy Fusion for Enhancing Localization in Wireless Sensor Networks (WSNs)

Localization in wireless sensor networks (WSNs) plays a crucial role in various applications thatrely on spatial information. This paper introduces the Multi-Strategy Fusion for Localization model, whichintegrates optimization techniques (ABO, DSA, EHO, and KNN) and neurocomputing techniques (BP, MTLSTM,BILSTM, and Autoencoder) to enhance localization accuracy in WSNs. The work is divided into three phases: datacollection, model building, and implementation. The first and the last are carried out in the field, while the secondis made in the laboratory. The three phases involve a few general steps. The (1) Data Collection Phase includesfour steps: (a) Deploy three anchors at known locations, forming an equilateral triangle. (b) Each anchor startsbroadcasting its location. (c) Using an ordinary sensor, the RSSI of each anchor is measured at every possiblelocation where the signal of the three anchors can reach it. (d) Data is logged to a CSV file containing themeasuring location and the RSSI of the three anchors and their locations. The Model Building Phase includes (a)preprocessing of the collected data, and (b) building a model based on optimization and optimization techniques.The Implementation phase includes five steps: (a) Convoy sensors to the target field. (b) Manually deploy anchorsaccording to the distribution plan. (c) Randomly deploy ordinary sensors. (d) Each ordinary sensor starts ininitialization mode. When receiving a signal from three anchors, a sensor computes its location and stores it forfuture use. When a sensor has its location, it turns into operational mode. (e) A sensor in the operational modeattaches the location with sensed data each time it sends it to the sink or neighbors according to routing protocols(routing is not considered in this study). ABO and DSA optimization techniques show similar performance, withlower Mean Squared Error (MSE) values compared to EHO and KNN. ABO and DSA also have similar MeanAbsolute Error (MAE) values, indicating lesser average absolute errors. BP emerges as the top performer amongthe neurocomputing techniques, demonstrating better accuracy with lower MSE and MAE values compared toMTLSTM, BILSTM, and Autoencoder. Finally; The Multi-Strategy Fusion for Localization model offers aneffective approach to enhance localization accuracy in wireless sensor networks. The paper focuses on addressingthe correlation between wireless device positions and signal intensities to improve the localization process. Theobtained results and provided justification emphasize the significance and value of the model in the field oflocalization in WSNs. The model represents a valuable contribution to the development of localization techniquesand improving their accuracy to meet the needs of various applications. The model opens up opportunities for itsutilization in diverse domains such as environmental monitoring, healthcare, smart cities, and disastermanagement, enhancing its practical applications and practical significance.

A collaboratively produced model of service design for children and young people with common mental health problems

BackgroundLittle is known about the effectiveness of, and implementation complexities associated with, service delivery models for children and young people (CYP) experiencing ‘common’ mental health problems such as anxiety, depression, behavioural difficulties and self-harm. This paper outlines how a model for high-quality service design for this population group was developed by identifying available services, their effectiveness, cost-effectiveness and acceptability, and the barriers and enablers to access.MethodsSequential, mixed-methods design, combining evidence syntheses (scoping and integrative reviews of the international literature) with primary research (a collective case study in England and Wales). Data from these two elements were collaboratively synthesised in a subsequent model-building phase.ResultsThe scoping review yielded a service model typology. The integrative review found effectiveness evidence only for four models: collaborative care (the only service model to also have cost-effectiveness evidence), outreach approaches, brief intervention services and an organisational framework called ‘Availability, Responsiveness and Continuity’. No service model seemed more acceptable than others. Three case study themes were identified: pathways to support; service engagement; and learning and understanding. The model-building phase identified rapid access, learning self-care skills, individualised support, clear information, compassionate and competent staff and aftercare planning as core characteristics of high-quality services. These characteristics were underpinned by four organisational qualities: values that respect confidentiality; engagement and involvement; collaborative relationships; and a learning culture.ConclusionsA consistent organisational evidence-base for service design and delivery in CYP’s mental health spanning many years appears to have had little impact on service provision in England and Wales. Rather than impose – often inflexible and untested – specific local or national models or frameworks, those commissioning, designing and delivering mental health services for CYP should (re)focus on already known, fundamental components necessary for high-quality services. These fundamental components have been integrated into a collaboratively produced general model of service design for CYP with common mental health problems. While this general model is primarily focused on British service provision, it is broad enough to have utility for international audiences.

Predicting Various Architectural Styles Using Computer Vision Methods

Computer Vision (CV), subfield of artificial intelligence (AI), enables computers to process visual data and recognize objects. CV is widely used in, automotive, food industry and diseases diagnosis. AI achieves this by algorithms. One of the important algorithms based on object detection is YOLO (You Only Look Once), provides more accurate results with high processing speed. The aim of this study is to perform an object detection-based CV project, to determine the structures in given video belong to one of the architectural styles: Gothic, Baroque, Palladian, or Art Nouveau. The study consists of data set creation, data labeling, model creation and model training. Roboflow was used as the data labeling platform and YOLOv8 was used for model building and training phases. At the end of the process, the fact that the model predicts architectural styles with high accuracy in a short time revealed that the model is a successful real-time object detection algorithm, and it was emphasized that CV can be used in the field of architecture and can contribute to other fields related to architecture.

Sample size and predictive performance of machine learning methods with survival data: A simulation study.

Prediction models are increasingly developed and used in diagnostic and prognostic studies, where the use of machine learning (ML) methods is becoming more and more popular over traditional regression techniques. For survival outcomes the Cox proportional hazards model is generally used and it has been proven to achieve good prediction performances with few strong covariates. The possibility to improve the model performance by including nonlinearities, covariate interactions and time-varying effects while controlling for overfitting must be carefully considered during the model building phase. On the other hand, ML techniques are able to learn complexities from data at the cost of hyper-parameter tuning and interpretability. One aspect of special interest is the sample size needed for developing a survival prediction model. While there is guidance when using traditional statistical models, the same does not apply when using ML techniques. This work develops a time-to-event simulation framework to evaluate performances of Cox regression compared, among others, to tuned random survival forest, gradient boosting, and neural networks at varying sample sizes. Simulations were based on replications of subjects from publicly available databases, where event times were simulated according to a Cox model with nonlinearities on continuous variables and time-varying effects and on the SEER registry data.

Bank Marketing Data Classification Using Optimized Voting Ensemble, Sine Cosine, and Genetic Algorithms

Nowadays, the banking industry is no exception to the general trend of massive data production in all spheres of modern life. In this research, we analyze the categorization of marketing data from banks using a variety of machine learning techniques. The term banking refers to the supply of services by a bank to an individual consumer. The data was first compiled from the UCI Machine Learning repository and the Kaggle website. Phone-based banking marketing statistics are the focus of this data set. Python is utilized as the language of implementation, and the Machine Learning concept is employed for statistical learning and data analysis in this work. An improved prediction is the primary goal of machine learning's model-building phase. In order to classify the results, a supervised Naive Bayes algorithm is used to the data. The primary goal of the modeling effort is to characterize whether or not the consumer has chosen a term deposit. The bank should devote substantial time to returning phone calls from prospective customers. Accuracy, precision, recall, and F1 score were all evaluated as a consequence of this study in the direction of term deposit forecasting.

Pose Estimation-Assisted Dance Tracking System Based on Convolutional Neural Network.

In the field of music-driven, computer-assisted dance movement generation, traditional music movement adaptations and statistical mapping models have the following problems: Firstly, the dance sequences generated by the model are not powerful enough to fit the music itself. Secondly, the integrity of the dance movements produced is not sufficient. Thirdly, it is necessary to improve the suppleness and rationality of long-term dance sequences. Fourthly, traditional models cannot produce new dance movements. How to create smooth and complete dance gesture sequences after music is a problem that needs to be investigated in this paper. To address these problems, we design a deep learning dance generation algorithm to extract the association between sound and movement characteristics. During the feature extraction phase, rhythmic features extracted from music and audio beat features are used as musical features, and coordinates of the main points of human bones extracted from dance videos are used for training as movement characteristics. During the model building phase, the model's generator module is used to achieve a basic mapping of music and dance movements and to generate gentle dance gestures. The identification module is used to achieve consistency between dance and music. The self-encoder module is used to make the audio function more representative. Experimental results on the DeepFashion dataset show that the generated model can synthesize the new view of the target person in any human posture of a given posture, complete the transformation of different postures of the same person, and retain the external features and clothing textures of the target person. Using a whole-to-detail generation strategy can improve the final video composition. For the problem of incoherent character movements in video synthesis, we propose to optimize the character movements by using a generative adversarial network, specifically by inserting generated motion compensation frames into the incoherent movement sequences to improve the smoothness of the synthesized video.

Detection of correct pregnancy status in lactating dairy cattle using MARS data mining algorithm

In this study, it is aimed to determine pregnancy outcomes by using multivariate adaptive regression splines (MARS) algorithm for classification type problems. For this purpose, data obtained from a private dairy farm in the Konya region of Türkiye in 2020 were used to determine pregnancy outcomes in Holstein dairy cattle. It has been determined how to perform statistical analyses on solving classification-type problems with the MARS algorithm and how to use R packages (caret and earth) by creating an R script file. After the analysis, the MARS estimation equation was created and in finding the probability of being pregnant: While lactation period, cow age, number of lactations, insemination number, and total lactation milk yield variables are important, it was seen that 7-day mean milk yield and last lactation milk yield were not significant. Using the train function of the caret package, the number of terms that produce the highest accuracy and the degree of interaction are determined. Goodness-of-fit tests of the optimum model were calculated. Within the scope of the evaluation of the generalization ability of the model, training and test sets were created, the classification success graph of the MARS algorithm, the model building phase were summarized, and the generalization ability of the established model was measured. When the pregnancy status is taken as a positive reference, the correct classification rate (sensitivity) of the animal with positive pregnancy status was found to be 0.9574. The correct classification rate (specificity) of pregnant animals was found to be 0.8370. The overall classification ratio of the training set (accuracy) was found to be 0.8777. The area under the ROC curve (AUC) was found to be 0.947, which indicates that the optimum specificity value is close to 1.

Space–time geological model of the Quaternary syntectonic fill of a foreland basin (Po basin, Northern Italy)

Quantitative models of large and widely populated Quaternary foreland basins, which host relevant resources like groundwater, heat, and fertile soils, may be computed if space–time models of the basin fills are available. The paper presents the combination of geometrical (space) and evolutionary (time) constraints to model Quaternary stratigraphy in the southern Po basin in Lombardy, disentangling the tectonic and climatic controls. Five Quaternary high-rank unconformities and seven intermediate-rank unconformities were recognised, down-traced, and correlated from surface to subsurface. Within this framework, the surface facies associations of high-, intermediate-, and low-rank stratigraphic units were compared to the litho-textural associations of their subsurface equivalents obtained from borehole logs and published geophysical images, to draw the best fitting model, which was constrained to the Quaternary geological evolution.Discrete, long-term thrusting, wrenching, folding, and extensional collapsing stages are inferred to have shaped the Gelasian, intra-Calabrian, Early–Middle Pleistocene, Middle–Late Pleistocene and Latest Pleistocene–Holocene high-rank composite unconformities. These surfaces are cut into synsedimentary, polyphasic, buried anticlines and correspond to conformable boundaries in the adjacent depocentres. The intermediate-rank, low-angle unconformities composing the high-rank ones, bound the stratigraphic units forming the overall regressive fill of the basin. These surfaces might correspond to steady tectonic uplift/subsidence periods when deposition was mostly controlled by Pleistocene climate changes since the Gelasian.The hierarchic space/time structure, translated into hard geometrical constraints (a fence diagram of cross-sections) and relative chronology of the stratigraphic relationships (intersect/erode, onlap and offset) allows computing the geological quantitative model using a potential field interpolation method. The model building phase and subsequent inspection permit critically analysing the current interpretations on the evolution of the Po basin during the Quaternary compared with other foreland settings. The 4-D model integrates the surface–subsurface geometry with the tectono-sedimentary evolution in a multi-scale structure that may be exported for hydrogeological, geothermal and environmental modelling.

Household-Level Energy Forecasting in Smart Buildings Using a Novel Hybrid Deep Learning Model

Forecasting of energy consumption in Smart Buildings (SB) and using the extracted information to plan and operate power generation are crucial elements of the Smart Grid (SG) energy management. Prediction of electrical loads and scheduling the generation resources to match the demand enable the utility to mitigate the energy generation cost. Different methodologies have been employed to predict energy consumption at different levels of distribution and transmission systems. In this paper, a novel hybrid deep learning model is proposed to predict energy consumption in smart buildings. The proposed framework consists of two stages, namely, data cleaning, and model building. The data cleaning phase applies pre-processing techniques to the raw data and adds additional features of lag values. In the model-building phase, the hybrid model is trained on the processed data. The hybrid deep learning (DL) model is based on the stacking of fully connected layers, and unidirectional Long Short Term Memory (LSTMs) on bi-directional LSTMs. The proposed model is designed to capture the temporal dependencies of energy consumption on dependent features and to be effective in terms of computational complexity, training time, and forecasting accuracy. The proposed model is evaluated on two benchmark energy consumption datasets yielding superior performance in terms of accuracy when compared with widely used hybrid models such as Convolutional (Conv) Neural Network-LSTM, ConvLSTM, LSTM encoder-decoder model, stacking models, etc. A mean absolute percentage error (MAPE) of 2.00% for case study 1 and a MAPE of 3.71% for case study 2 is obtained for the proposed forecasting DL model in comparison with LSTM-based models that yielded 7.80% MAPE and 5.099% MAPE for two datasets respectively. The proposed model has also been applied for multi-step week-ahead daily forecasting with an improvement of 8.368% and 20.99% in MAPE against the LSTM-based model for the utilized energy consumption datasets respectively.

An innovative monitoring and maintenance model for the INFN CNAF Tier-1 data center infrastructure

During the last years we have carried out a renewal of the Building Management System (BMS) software of our data center with the aim of improving the data collection capability. Considering the complex physical distribution of the technical plants and the limits of the actual building hosting our center, a system that simply monitors and collects all the necessary information and provides alarms only in case of major failures has proven to be unsatisfactory. In 2017 we suffered a major flood due to one main water pipeline failure in the public street. After this disastrous event, clearly far beyond our control, we were however forced to reconsider completely the physical site robustness of our building in addition to the current monitoring and alarm system capabilities. It was clear that in some specific cases, alerts should be triggered hours or days before the actual main problem arises in order to allow efficient human intervention and proper escalation process. This paradigm could be easily applied to almost all the infrastructure components in our site, mainly the electric power distribution and continuity systems as well as the whole cooling devices. For this reason, in parallel to a consistent increase in the sensor widespread distribution of our BMS data collector system, a study of a predictive maintenance approach applicability to our site has been started. Predictive maintenance techniques aims at prevent unexpected infrastructure components failures or major events with the study of the whole monitoring data collection and the creation of appropriate statistical models with the help of big data analysis and machine learning techniques. An improvement in the Power Distribution Units (PDUs) monitoring in our site and the introduction of a dedicated network of water leak sensors were the first steps for increasing the data collection information at our disposal. With sufficient monitoring statistical information stored in our BMS system a preliminary and exploratory predictive data analysis proof of concept could be constructed. This could lead to the model building phase and the creation of a prototype with the aim of forecasting future infrastructure main failure events and forthcoming error conditions. The general idea is, conceivably, an approach to the predictive maintenance model where it would be possible to introduce scheduled corrective actions for the purpose of preventing potential failures in the next future and increasing the site overall reliability.

Automated Framework for Time-Domain Piecewise-Linear Fitting Method Based on Digital Wave Processing of $S$ -Parameters

An innovative full time-domain macromodeling technique for general, linear multiport systems is described. The methodology is defined in a digital wave framework and time-domain simulations are performed via an efficient method called Segment Fast Convolution (SFC). It is based on a piecewise-constant (PWC) model of the impulse response of scattering parameters, computed starting from a piecewise-linear fitting of their step response (PWLFIT). Such step response is directly available from time-domain reflectometer measurements (TDR/TDT) or equivalent simulations. The model-building phase is performed in a fast automated framework and an analytic formulation of computational efficiency of the SFC with respect to the standard time-domain convolution is given. Two application examples are used to verify the PWLFIT performance and to perform a comparison with macromodeling methods defined in the frequency-domain, such as Vector Fitting (VF).

Revitalizing seismic data with new imaging solutions to positively impact field development

Evaluating, planning, and forecasting are integral parts of asset development and continue throughout the life cycle of a producing field. The right decisions are required to lower risk and maximize economic recovery in challenging environments. The Claymore Complex is located in the North Sea and was discovered in 1977. A number of geologic challenges affect the imaging and hence field development including a system of shallow interweaving Quaternary channels, numerous high-contrast layers of varying composition, overburden structural complexity, and a sequence of tilted fault blocks containing the main reservoir systems. Historically, seismic processing over the area has not fully solved these challenges, resulting in significant imaging uncertainty. The Claymore Complex has an abundance of data including a large population of well information and interpretation. As part of a data revitalization process, geostatistical integration of these auxiliary data into a velocity model building sequence using full-waveform inversion and wavelet shift tomography enabled the generation of an accurate high-resolution velocity model. Access to a recent 3D survey acquired obliquely to existing data improved subsurface illumination for both the model building and imaging phases. Near-surface imaging effects and their impact on reservoir positioning and clarity were improved using the upgraded velocity model and dual-azimuth data. Shallow imaging challenges were mitigated by utilizing the additional illumination and angular diversity contained within the multiple reverberations. The revitalization of the Claymore area seismic data has challenged the current understanding of the geologic framework. Confidence has been improved by solving depth conversion problems and increasing the understanding of fault positioning and reservoir connectivity, which are invaluable for future field development.

Healthcare decision‐making applications using multicriteria decision analysis: A scoping review

AbstractThe objective of this study is to review applications of multicriteria decision analysis (MCDA) methods to provide structure for and practical insights on how MCDA methods are applied in different healthcare areas. Four databases (Embase, Medline, Web of Science, and PubMed) were searched from 1980 to February 7, 2018. A scoping review was performed to outline the most frequently applied MCDA methods, in which only case studies were considered, as the objective was to assess the applications of MCDA methods. Seventy case studies were selected. The review showed that the first real‐life application was published in 1990. With regard to areas of application, three aspects were analysed: type of health services (public or private), type of intervention (preventive, diagnostic, or treatment), and healthcare area (policymaking; resource allocation; health technology assessment; pharmaceutics; and hospital, clinical, or patient level). Data revealed that MCDA was principally used in public health services (n = 59). Three types of interventions were identified: prevention (n = 36), diagnosis (n = 32), and treatment (n = 2). An MCDA process of application was then followed to provide practical insights. During the problem‐structuring phase, results showed that processes for selecting stakeholders and criteria differed among healthcare areas. For model building, authors preferred value measurement methods and especially the analytical hierarchy process. Authors of the 70 case studies included and confirmed the usefulness of MCDA in different healthcare contexts. The stakeholders, criteria, and MCDA methods chosen differ from one study to another, answering different research objectives in different healthcare areas. This scoping review highlights the importance of each MCDA step (using Belton and Stewart's framework), from problem structuring (including the selection of stakeholders) to the model building phase (encompassing the selection of the appropriate MCDA method). We recommend structured and justified analysis so that decision makers and stakeholders can feel confident during the application process in order to make final decisions more meaningful. Further work is needed to help researchers and decision makers when choosing a suitable MCDA method.

Connected and autonomous vehicles: A cyber-risk classification framework

The proliferation of technologies embedded in connected and autonomous vehicles (CAVs) increases the potential of cyber-attacks. The communication systems between vehicles and infrastructure present remote attack access for malicious hackers to exploit system vulnerabilities. Increased connectivity combined with autonomous driving functions pose a considerable threat to the vast socioeconomic benefits promised by CAVs. However, the absence of historical information on cyber-attacks mean that traditional risk assessment methods are rendered ineffective. This paper proposes a proactive CAV cyber-risk classification model which overcomes this issue by incorporating known software vulnerabilities contained within the US National Vulnerability Database into model building and testing phases. This method uses a Bayesian Network (BN) model, premised on the variables and causal relationships derived from the Common Vulnerability Scoring Scheme (CVSS), to represent the probabilistic structure and parameterisation of CAV cyber-risk. The resulting BN model is validated with an out-of-sample test demonstrating nearly 100% prediction accuracy of the quantitative risk score and qualitative risk level. The model is then applied to the use-case of GPS systems of a CAV with and without cryptographic authentication. In the use case, we demonstrate how the model can be used to predict the effect of risk reduction measures.

A novel case-control subsampling approach for rapid model exploration of large clustered binary data.

In many settings, an analysis goal is the identification of a factor, or set of factors associated with an event or outcome. Often, these associations are then used for inference and prediction. Unfortunately, in the big data era, the model building and exploration phases of analysis can be time-consuming, especially if constrained by computing power (ie, a typical corporate workstation). To speed up this model development, we propose a novel subsampling scheme to enable rapid model exploration of clustered binary data using flexible yet complex model set-ups (GLMMs with additive smoothing splines). By reframing the binary response prospective cohort study into a case-control-type design, and using our knowledge of sampling fractions, we show one can approximate the model estimates as would be calculated from a full cohort analysis. This idea is extended to derive cluster-specific sampling fractions and thereby incorporate cluster variation into an analysis. Importantly, we demonstrate that previously computationally prohibitive analyses can be conducted in a timely manner on a typical workstation. The approach is applied to analysing risk factors associated with adverse reactions relating to blood donation.

Multiple criteria decision analysis in the context of health technology assessment: a simulation exercise on metastatic colorectal cancer with multiple stakeholders in the English setting

BackgroundMultiple criteria decision analysis (MCDA) has appeared as a methodology to address limitations of economic evaluation in health technology assessment (HTA), however there are limited empirical evidence from real world applications. The aim of this study is to test in practice a recently developed MCDA methodological framework known as Advance Value Framework (AVF) through a proof-of-concept case study engaging multiple stakeholders.MethodsA multi-attribute value theory methodological process was adopted involving problem structuring, model building, model assessment and model appraisal phases. A facilitated decision analysis modelling approach was used as part of a decision conference with thirteen participants. An expanded scope of the National Institute for Health and Care Excellence (NICE) remit acted as the study setting with the use of supplementary value concerns. Second-line biological treatments were evaluated for metastatic colorectal cancer (mCRC) patients having received prior chemotherapy, including cetuximab monotherapy, panitumumab monotherapy and aflibercept in combination with FOLFIRI chemotherapy. Initially 18 criteria attributes were considered spanning four value domains relating to therapeutic impact, safety profile, innovation level and socioeconomic impact.ResultsNine criteria attributes were finally included. Cetuximab scored the highest overall weighted preference value score of 45.7 out of 100, followed by panitumumab with 42.3, and aflibercept plus FOLFIRI with 14.4. The relative weights of the two most important criteria (overall survival and Grade 4 adverse events) added up to more than the relative weight of all other criteria together (52.1%). Main methodological limitation was the lack of comparative clinical effects across treatments and challenges included the selection of “lower” and “higher” reference levels on criteria attributes, eliciting preferences across attributes where participants had less experience, and ensuring that all attributes possess the right decision theory properties.ConclusionsThis first application of AVF produced transparent rankings for three mCRC treatments based on their value, by assessing an explicit set of evaluation criteria while allowing for the elicitation and construction of participants’ value preferences and their trade-offs. It proved it can aid the evaluation process and value communication of the alternative treatments for the group participants. Further research is needed to optimise its use as part of policy-making.