Sensitivity Analysis Of Inputs Research Articles

Optimising HIV drug resistance testing laboratory networks in Kenya: insights from systems engineering modelling

BackgroundHIV drug resistance (DR) is a growing threat to the durability of current and future HIV treatment success. DR testing (DRT) technologies are very expensive and specialised, relying on centralised...

Sensitivity Analysis and Uncertainty Quantification of Nanoparticle Deposition from Tongue Morphological Variations.

The human tongue has highly variable morphology. Its role in regulating respiratory flows and deposition of inhaled aerosols remains unclear. The objective of this study was to quantify the uncertainty of nanoparticle deposition from the variability in tongue shapes and positions and to rank the importance of these morphological factors. Oropharyngeal models with different tongue postures were reconstructed by modifying an existent anatomically accurate upper airway geometry. An LRN k-ω model was applied to solve the multiregime flows, and the Lagrangian tracking approach with near-wall treatment was used to simulate the behavior and fate of inhaled aerosols. Once the database of deposition rates was completed, a surrogate model was trained using Gaussian process regression with polynomial kernels and was validated by comparing its predictions to new CFD simulations. Input sensitivity analysis and output updateability quantification were then performed using the surrogate model. Results show that particle size is the most significant parameter in determining nanoparticle deposition in the upper airway. Among the morphological factors, the shape variations in the central tongue had a higher impact on the total deposition than those in the back tongue and glottal aperture. When considering subregional deposition, mixed sensitivity levels were observed among morphological factors, with the back tongue being the major factor for throat deposition and the central tongue for oral deposition. Interaction effects between flow rate and morphological factors were much higher than the effects from individual parameters and were most significant in the throat (pharyngolaryngeal region). Given input normal variances, the nanoparticle deposition exhibits logarithmical normal distributions, with much lower uncertainty in 100-nm than 2-nm aerosols.

TemproNet: A transformer-based deep learning model for seawater temperature prediction

Accurate prediction of seawater temperature is crucial for meteorological model understanding and climate change assessment. This study proposes TempreNet, a deep learning model based on a transformer and convolutional neural network, to accurately predict subsurface seawater temperature using satellite observations in the South China Sea. TemproNet uses multivariate sea surface observations such as sea level anomaly (SLA), sea surface temperature (SST), and sea surface wind (SSW) as model inputs, which employs a hierarchical transformer encoder to extract the multi-scale feature, uses a lightweight convolutional decoder to predict seawater temperature. We train and validate the model using the CMEMS temperature dataset and compare its accuracy with Attention-Unet, LightGMB, and ANN. Experimental results show that TemproNet has significantly outperformed other models with RMSE and R2 of 0.52 °C and 0.83 in a 32-layer temperature profile prediction task over 200 m in the South China Sea. In addition, we fully demonstrate the error of our model in space, in time, and at different depths, showing the efficiency and stability of our model. The input sensitivity analysis showed that SST contributed more to predicting shallow water temperature, while SLA significantly impacted the prediction of mid-deep water temperature. The results of this study provide an innovative and reliable solution for seawater temperature prediction and have important implications for meteorological model understanding and climate change assessment.

An agent-based model of cardiac allograft vasculopathy: toward a better understanding of chronic rejection dynamics.

Cardiac allograft vasculopathy (CAV) is a coronary artery disease affecting 50% of heart transplant (HTx) recipients, and it is the major cause of graft loss. CAV is driven by the interplay of immunological and non-immunological factors, setting off a cascade of events promoting endothelial damage and vascular dysfunction. The etiology and evolution of tissue pathology are largely unknown, making disease management challenging. So far, in vivo models, mostly mouse-based, have been widely used to study CAV, but they are resource-consuming, pose many ethical issues, and allow limited investigation of time points and important biomechanical measurements. Recently, agent-based models (ABMs) proved to be valid computational tools for deciphering mechanobiological mechanisms driving vascular adaptation processes at the cell/tissue level, augmenting cost-effective in vivo lab-based experiments, at the same time guaranteeing richness in observation time points and low consumption of resources. We hypothesize that integrating ABMs with lab-based experiments can aid in vivo research by overcoming those limitations. Accordingly, this work proposes a bidimensional ABM of CAV in a mouse coronary artery cross-section, simulating the arterial wall response to two distinct stimuli: inflammation and hemodynamic disturbances, the latter considered in terms of low wall shear stress (WSS). These stimuli trigger i) inflammatory cell activation and ii) exacerbated vascular cell activities. Moreover, an extensive analysis was performed to investigate the ABM sensitivity to the driving parameters and inputs and gain insights into the ABM working mechanisms. The ABM was able to effectively replicate a 4-week CAV initiation and progression, characterized by lumen area decrease due to progressive intimal thickening in regions exposed to high inflammation and low WSS. Moreover, the parameter and input sensitivity analysis highlighted that the inflammatory-related events rather than the WSS predominantly drive CAV, corroborating the inflammatory nature of the vasculopathy. The proof-of-concept model proposed herein demonstrated its potential in deepening the pathology knowledge and supporting the in vivo analysis of CAV.

Recommendations for finite element modelling of nickel-titanium stents-Verification and validation activities.

The objective of this study is to present a credibility assessment of finite element modelling of self-expanding nickel-titanium (Ni-Ti) stents through verification and validation (VV) activities, as set out in the ASME VV-40 standard. As part of the study, the role of calculation verification, model input sensitivity, and model validation is examined across three different application contexts (radial compression, stent deployment in a vessel, fatigue estimation). A commercially available self-expanding Ni-Ti stent was modelled, and calculation verification activities addressed the effects of mesh density, element integration and stable time increment on different quantities of interests, for each context of use considered. Sensitivity analysis of the geometrical and material input parameters and validation of deployment configuration with in vitro comparators were investigated. Results showed similar trends for global and local outputs across the contexts of use in response to the selection of discretization parameters, although with varying sensitivities. Mesh discretisation showed substantial variability for less than 4 × 4 element density across the strut cross-section in radial compression and deployment cases, while a finer grid was deemed necessary in fatigue estimation for reliable predictions of strain/stress. Element formulation also led to substantial variation depending on the chosen integration options. Furthermore, for explicit analyses, model results were highly sensitive to the chosen target time increment (e.g., mass scaling parameters), irrespective of whether quasistatic conditions were ensured (ratios of kinetic and internal energies below 5%). The higher variability was found for fatigue life simulation, with the estimation of fatigue safety factor varying up to an order of magnitude depending on the selection of discretization parameters. Model input sensitivity analysis highlighted that the predictions of outputs such as radial force and stresses showed relatively low sensitivity to Ni-Ti material parameters, which suggests that the calibration approaches used in the literature to date appear reasonable, but a higher sensitivity to stent geometry, namely strut thickness and width, was found. In contrast, the prediction of vessel diameter following deployment was least sensitive to numerical parameters, and its validation with in vitro comparators offered a simple and accurate (error ~ 1-2%) method when predicting diameter gain, and lumen area, provided that the material of the vessel is appropriately characterized and modelled.

Economic Aspects of Profit Maximization if Cost of Principal Raw Material Increases

In this study sensitivity analysis of various inputs is discussed, when per unit cost of principal raw materials is increased. The organization has its main target of achieving maximum profit through the adjustment of various inputs and outputs in future production. In this study Cobb-Douglas production function is considered as a profit function to investigate sensitivity analysis during profit maximization procedures. In this article the method of Lagrange multiplier is applied to represent higher dimensional unconstrained problem from the lower dimensional constrained problem. Moreover, the determinant of the 6×6 bordered Hessian matrix and 6×6 Jacobian are applied for the augmentation of the sensitivity analysis efficiently.

Sensitivity Analysis of Inputs of an Organization: A Profit Maximization Exploration

This article tries to discuss sensitivity analysis of various inputs of an organization during profit maximization investigations. In this study Cobb-Douglas production function is analyzed with a detail mathematical analysis. The method of Lagrange multiplier is a very useful and powerful technique in multivariate calculus, and it is applied here to obtain higher dimensional unconstrained problem from the lower dimensional constrained one. In the article determinant of the 6×6 bordered Hessian matrix and 6×6 Jacobian are operated for consulting sensitivity analysis efficiently.

An Application of Machine Learning to Estimate and Evaluate the Energy Consumption in an Office Room

There are no exact criteria for the architecture of openings and windows in office buildings in order to optimize energy consumption. Due to the physical limitations of this renewable energy source and the lack of conscious control over its capabilities, the amount of light entering offices and the role of daylight as a source of energy are determined by how they are constructed. In this study, the standard room dimensions, which are suitable for three to five employees, are compared to computer simulations. DesignBuilder and EnergyPlus are utilized to simulate the office’s lighting and energy consumption. This study presents a new method for estimating conventional energy consumption based on gene expression programming (GEP). A gravitational search algorithm (GSA) is implemented in order to optimize the model results. Using input and output data collected from a simulation of conventional energy use, the physical law underlying the problem and the relationship between inputs and outputs are identified. This method has the advantages of being quick and accurate, with no simulation required. Based on effective input parameters and sensitivity analysis, four models are evaluated. These models are used to evaluate the performance of the trained network based on statistical indicators. Among all the GEP models tested in this study, the one with the lowest MAE (0.1812) and RMSE (0.09146) and the highest correlation coefficient (0.90825) is found to be the most accurate.

Machine learning-based automated classification of worker-reported safety reports in construction

Limited academic attention has been paid to the applicability of Machine Learning (ML) approaches for analyzing worker-reported near-miss safety reports, as opposed to injury reports, at construction sites. Although resource-efficient analysis through ML of large volumes of such data at construction sites can help guide practitioners in decision-making to prevent injuries. The current study addresses this research gap by evaluating the relevance of ML approaches through quantitative and qualitative methods for scaling efficient near-miss reporting programs at construction sites. The study uses an extensive experimentation strategy consisting of input data processing, n-gram modeling, and sensitivity analysis. It first tests the proposition that, despite the data-quality challenges, the high performance of different ML algorithms can be achieved in automatically classifying the textual near-miss observations. The study relies on worker-reported near-miss data collected from a real construction site in Kuwait. The classification performance of various ML approaches is evaluated using F1 scores for three academically novel but commonly used category labels at the sites - "Unsafe Act (UA)," "Unsafe Condition (UC)," and "Good Observation (GO)." In addition, the practitioner's input was utilized to assess the practical applicability of ML classifiers for construction sites. The conventional Logistic Regression (LR) classifiers have a comparatively high F1 score of 0.79. However, ML classifiers faced challenges in distinguishing between UA and UC. Further, the analysis reveals that optimal ML classifiers may lose on being acceptable to human decision-makers. Overall, despite the promising performance of ML tools for the near-miss data, the sites with low maturity of reporting systems may find themselves unable to leverage ML to scale their reporting systems. A simplified experimentation strategy like the current study could help practitioners identify the data-specific optimal ML approaches in future applications.

Safety assessment using computer experiments and surrogate modeling: Railway vehicle safety and track quality indices

Mathematical modeling and advances in computation allow exploring multiple scenarios and studying the reliability and safety of transportation systems. Although track geometry directly impacts vehicle safety, the track quality indices used by infrastructure managers to assess tracks seldom consider vehicle dynamics. This work provides a design and analysis of computer experiments framework to model the relationships between track quality and vehicle safety. The framework considers input selection and pre-processing, vehicle responses and post-processing, input screening, surrogate modeling, sensitivity analysis, and safety assessment. This approach allows studying how track geometry parameters and other variables influence safety quantities. The framework is demonstrated with a case study that combines two European standards: the standard for track geometry quality, EN 13848, and the standard for vehicle acceptance, EN 14363. The case study considers different vehicle types, vehicle speed, track curvature, track flexibility, and track irregularities. The results show, for each safety quantity, which inputs are relevant. In particular, the sensitivity analysis indicates two influential inputs not considered in EN 13848 that could help assess track condition. Finally, an example illustrates how these surrogates can be used to find which safety quantities govern safety and define track geometry limits directly linked to vehicle safety.

Costs and benefits of routine hemoglobin A1c screening prior to total joint arthroplasty: a cost-benefit analysis.

Poorly controlled diabetes mellitus (DM) increases the risk for periprosthetic joint infection (PJI) after total joint arthroplasty (TJA). While institutional protocols include hemoglobin A1c (HbA1c) screening in TJA patients, the costs and benefits of routine preoperative screening have not been described. The authors created a decision tree model to evaluate short-term costs and risk reduction for PJIs with routine screening of primary total hip arthroplasty (THA) and total knee arthroplasty (TKA) patients. Probabilities and costs were obtained from published sources. They calculated net costs and absolute risk reduction in PJI for routine screening versus no screening. The authors also performed sensitivity analyses of model inputs including probabilistic sensitivity analyses (PSAs) consisting of 10,000 Monte Carlo simulations. In patients with DM, routine screening before THA resulted in net cost savings of $81 per patient with 286 patients needing to be screened to prevent 1 PJI, while screening before TKA incurred net additional costs of $25,810 per PJI prevented. Routine screening in patients with DM undergoing THA or TKA was cost-saving across 75.5% or 21.8% of PSA simulations, respectively. In patients with no history of DM, routine screening before THA or TKA incurred net additional costs of $24,583 or $87,873 per PJI prevented, respectively. Routine HbA1c screening in patients with DM prior to THA with referral of patients with elevated HbA1c for glycemic optimization may prevent PJI and reduce healthcare costs. In contrast, routine screening in patients with DM prior to TKA or in patients with no history of DM is not cost-saving. Economic Level IV.

Batch-sequential design and heteroskedastic surrogate modeling for delta smelt conservation

Delta smelt is an endangered fish species in the San Francisco estuary that have shown an overall population decline over the past 30 years. Researchers have developed a stochastic, agent-based simulator to virtualize the system with the goal of understanding the relative contribution of natural and anthropogenic factors that might play a role in their decline. However, the input configuration space is high dimensional, running the simulator is time-consuming, and its noisy outputs change nonlinearly in both mean and variance. Getting enough runs to effectively learn input–output dynamics requires both a nimble modeling strategy and parallel evaluation. Recent advances in heteroskedastic Gaussian process (HetGP) surrogate modeling helps, but little is known about how to appropriately plan experiments for highly distributed simulation. We propose a batch sequential design scheme, generalizing one-at-a-time variance-based active learning for HetGP, as a means of keeping multicore cluster nodes fully engaged with runs. Our acquisition strategy is carefully engineered to favor selection of replicates which boost statistical and computational efficiency when training surrogates to isolate signal from noise. Design and modeling are illustrated on a range of toy examples before embarking on a large-scale smelt simulation campaign and downstream high-fidelity input sensitivity analysis.

Karst spring discharge modeling based on deep learning using spatially distributed input data

Abstract. Despite many existing approaches, modeling karst water resources remains challenging as conventional approaches usually heavily rely on distinct system knowledge. Artificial neural networks (ANNs), however, require only little prior knowledge to automatically establish an input–output relationship. For ANN modeling in karst, the temporal and spatial data availability is often an important constraint, as usually no or few climate stations are located within or near karst spring catchments. Hence, spatial coverage is often not satisfactory and can result in substantial uncertainties about the true conditions in the catchment, leading to lower model performance. To overcome these problems, we apply convolutional neural networks (CNNs) to simulate karst spring discharge and to directly learn from spatially distributed climate input data (combined 2D–1D CNNs). We investigate three karst spring catchments in the Alpine and Mediterranean region with different meteorological–hydrological characteristics and hydrodynamic system properties. We compare the proposed approach both to existing modeling studies in these regions and to our own 1D CNN models that are conventionally trained with climate station input data. Our results show that all the models are excellently suited to modeling karst spring discharge (NSE: 0.73–0.87, KGE: 0.63–0.86) and can compete with the simulation results of existing approaches in the respective areas. The 2D models show a better fit than the 1D models in two of three cases and automatically learn to focus on the relevant areas of the input domain. By performing a spatial input sensitivity analysis, we can further show their usefulness in localizing the position of karst catchments.

Modeling the health and economic implications of adopting a 1-dose 9-valent human papillomavirus vaccination regimen in a high-income country setting: An analysis in the United Kingdom

Although no human papillomavirus (HPV) vaccine is indicated for single-dose administration, some observational evidence suggests that a 1-dose regimen might reduce HPV infection risk to that achieved with 2 doses. This study estimated the potential health and economic outcomes associated with switching from a 2-dose HPV vaccination program for girls and boys aged 13–14 years to an off-label 9-valent (9vHPV), 1-dose regimen, accounting for the uncertainty of the effectiveness and durability of a single dose. A dynamic HPV transmission infection and disease model was adapted to the United Kingdom and included a probabilistic sensitivity analysis using estimated distributions for duration of protection of 1-dose and degree of protection of 1 relative to 2 doses. One-way sensitivity analyses of key inputs were performed. Outcomes included additional cancer and disease cases and the difference in net monetary benefit (NMB). The 1-dose program was predicted to result in 81,738 additional HPV-related cancer cases in males and females over 100 years compared to the 2-dose program, ranging from 36,673 to 134,347 additional cases (2.5% and 97.5% quantiles, respectively), and had a 7.8% probability of being cost-effective at the £20,000/quality-adjusted life years willingness-to-pay (WTP) threshold. In one-way sensitivity analyses, the number of additional cancer cases was sensitive to the median of the duration of protection distribution and coverage rates. The differences in NMBs were sensitive to the median of the duration of protection distribution, dose price and discount rate, but not coverage variations. Across sensitivity analyses, the probability of 1 dose being cost-effective vs 2 doses was < 50% at the standard WTP threshold. Adoption of a 1-dose 9vHPV vaccination program resulted in more vaccine-preventable HPV-related cancer and disease cases in males and females, introduced substantial uncertainty in health and economic outcomes, and had a low probability of being cost-effective compared to the 2-dose program.

Sensitivity analysis and optimisation of HVOF process inputs to reduce porosity and maximise hardness of WC-10Co-4Cr coatings

Sensitivity analysis and optimisation of HVOF process inputs to reduce porosity and maximise hardness of WC-10Co-4Cr coatings

Global Sensitivity Analysis for Canopy Reflectance and Vegetation Indices of Mangroves

Remote sensing has been applied to map the extent and biophysical properties of mangroves. However, the impact of several critical factors, such as the fractional cover and leaf-to-total area ratio of mangroves, on their canopy reflectance have rarely been reported. In this study, a systematic global sensitivity analysis was performed for mangroves based on a one-dimensional canopy reflectance model. Different scenarios such as sparse or dense canopies were set up to evaluate the impact of various biophysical and environmental factors, together with their ranges and probability distributions, on simulated canopy reflectance spectra and selected Sentinel-2A vegetation indices of mangroves. A variance-based method and a density-based method were adopted to compare the computed sensitivity indices. Our results showed that the fractional cover and leaf-to-total area ratio of mangrove crowns were among the most influential factors for all examined scenarios. As for other factors, plant area index and water depth were influential for sparse canopies while leaf biochemical properties and inclination angles were more influential for dense canopies. Therefore, these influential factors may need attention when mapping the biophysical properties of mangroves such as leaf area index. Moreover, a tailored sensitivity analysis is recommended for a specific mapping application as the computed sensitivity indices may be different if a specific input configuration and sensitivity analysis method are adopted.

TRIPLEX-Mortality model for simulating drought-induced tree mortality in boreal forests: Model development and evaluation

Globally, increasing drought-induced tree mortality rates under climate change are projected to have far-reaching effects on forest ecosystems. Among these forest systems, the boreal forest is considered a ‘tipping element’ of the Earth's climate system. This forest biome plays a critical role in ecosystem services, structures and functions while being highly sensitive to drought stress. Although process-based models are important tools in ecological research, very few have yet been developed that integrate advanced physiological mechanisms to simulate drought-induced mortality in boreal forests. Accordingly, based on the process-based TRIPLEX model, this study introduces the new TRIPLEX-Mortality submodule for the Canadian boreal forests at the stand level, that for the first time successfully incorporates two advanced drought-induced physiological mortality mechanisms (i.e., hydraulic failure and carbon starvation). To calibrate and validate the model, 73 permanent sample plots (PSPs) were selected across Canada's boreal forests. Results confirm a good agreement between simulated mortality and mortality observations (R2=0.79; P<0.01; IA=0.94), demonstrating good model performance in simulating drought-induced mortality in boreal forests. Sensitivity analysis indicated that parameter sensitivity increased as drought intensified, and the shape parameter (c) for calculating percentage loss of conductivity (PLC) was the most sensitive parameter (average SI = -3.51) to simulate tree mortality. Furthermore, the results of model input sensitivity analysis also showed that the model can capture changes in mortality under different drought scenarios. Consequently, our model is suitable for simulating drought-induced mortality in boreal forests while also providing new insight into improving model simulations for tree mortality and associated carbon dynamics in a progressively warmer and drier world.

Specific heat capacity of molten salt-based nanofluids in solar thermal applications: A paradigm of two modern ensemble machine learning methods

The quantitative determination of specific heat capacity (SHC) of molten (nitrate) salt-based nanofluids helps to control the start-up heat and prevent overheating when deployed as a working heat transfer fluid in a wide range of solar thermal applications. Thus, accurate measurement of the SHC and capturing the melting point is of paramount importance in the molten salt-based nanofluids’ characterization analyses applied in solar collectors. In this research, two modern ensemble machine learning models, Extra Tree Regression (ETR) and AdaBoost Regression (ABR), were developed based on 2,384 datasets, including solid mass fraction (w), temperature (T), SHC of base fluid (CPBase), mean diameter (Dp), and density (ρp) of nanoparticle as all independent input variables and the SHC of molten salt-based nanofluids (CPMS-nf) as the target. Herein, the stepwise forward method and mutual information were addressed to determine the best input combination and sensitivity analysis. The provided models were validated using Random Forest (RF) and Boosted Regression Tree (BRT) as two powerful other ensemble models. The results demonstrated that ETR model in terms of (R = 0.9964, RMSE = 0.1566, and U95%=3.6062) outperformed the ABR (R = 0.9949, RMSE = 0.1855, and U95%=3.6009), RF (R = 0.9922, RMSE = 0.2326, and U95%=3.5904), and BRT (R = 0.9907, RMSE = 0.2508, and U95%=3.5857). The SHC of molten salt base fluid was identified as the most significant factor in estimating the SHC of molten salt-based nanofluids.

Analysis and Evaluation of Energy, Economic and Environmental Impact Indicators of Horticultural and Greenhouse Production Systems in Iran using PRISMA Method

AbstractIntroductionSo far, many studies have been conducted to evaluate the impact of input consumption pattern on energy, economic and environmental indicators in horticultural and greenhouse crops in Iran. A review of these studies shows that the causes of the current situation in the systems have not been investigated. These studies are mostly reporting the current situation and the interventions and their effect on improving the input consumption pattern in the sustainability of the system have not been considered by researchers. Also, studies showed that the study location and products do not fit well with the volume of production in the horticultural and greenhouse sector of the Iran. Therefore, in order to increase the effectiveness and future direction of studies in this field, this review study was conducted. In this article, Iranian horticultural and greenhouse production systems were reviewed and analyzed by reviewing the published articles between 2008 and 2018, using the PRISMA method. The PRISMA method is a well-known method for conducting systematic review studies. The PRISMA method includes following sections: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings. In this article, 16 types of garden products and 6 types of greenhouse products were studied.Material and MethodsIn this study, the methods used to determine the status of energy consumption pattern, economic and environmental in horticultural and greenhouse crops were analyzed. For this purpose, the indicators of total energy consumption (TEI), energy efficiency (EUE), net energy (NE), energy efficiency (EP) were examined in the section of energy. The issue of sensitivity analysis of energy inputs was also examined and the highest values of t-statistic and MPP were reported for products. In some articles, data envelopment analysis method was used in systems performance analysis. The indicators used included technical efficiency (TE), pure technical efficiency (PTE), scale efficiency (SE) and energy saving target ratio (ESTR). The results of them were summarized and reported. In some studies, the method of artificial neural networks and Adaptive Neuro-Fuzzy Inference System were used. In general, in the present article, the challenges and risks in the methods used in previous studies were considered. The issue of sampling in the analysis of agricultural systems was discussed in detail and a new sampling procedure was proposed. To draw a general picture of energy and environmental indicators of orchard and greenhouse systems in Iran, the results published in the articles were reviewed. Not all researchers use the same equivalents in calculating the indices, and this makes the results of the studies slightly different from each other. The existence of such differences causes some deviations in comparing the results of similar articles in the same products. However, to adjust for these differences, averaging was used in the index report.Results and DiscussionThe study of the share of inputs in the total energy consumption shows that in horticultural products, the share of fertilizer and electricity inputs is very significant. In the case of greenhouse products, fuel input, which is mainly diesel, has the largest share of energy consumption. Walnuts have the lowest energy consumption and strawberries have the highest energy consumption among orchard products. Grapes, apples and walnuts also have positive net energy, so they have the highest energy efficiency compared to other products. The most important inputs that have the greatest potential for energy savings in most products are diesel fuel and electricity. Among greenhouse crops in cucumber production, diesel fuel has great potential for energy savings that need to be reduced in future research. In the case of strawberry and rose, electricity input has the greatest potential for energy savings. Knowing the potential of inputs that can be saved can be effective in changing the behavior of producers.ConclusionsTo increase the effectiveness of research in this area, such studies should be done dynamically and for at least two or more years. In the first year, the input consumption pattern should be extracted and after performing the consumption pattern modifying interventions, the effect of these actions should be evaluated in the following years. Data envelopment analysis methods and multi-objective genetic algorithm can be well used to develop solutions to improve input consumption pattern. The review of articles showed that the study of the effect of social factors on the behavior of various production systems has been neglected. Since the pattern of energy consumption in the agricultural sector is significantly dependent on the behavior of users and the characteristics of systems and methods of production, it seems necessary to pay attention to this factor to prepare and design any process improvement strategy in the system. In this study, a new procedure including three stages of analysis, redesign and evaluation was proposed to complete the studies related to the analysis of agricultural systems.

A Real Estate Valuation Model Using Boosted Feature Selection

To estimate real estate values, a complex valuation model based on artificial neural network (ANN) has been established as a successful means in modern machine learning research, specifically when high-dimensional data are available. Unfortunately, the real estate data in many locations, such as Thailand, are quite limited in terms of features. Hence, it becomes mandatory to reduce the complexity using feature selection techniques. These techniques aim to improve performance by identifying significant factors and help decrease the computational overload and model construction. However, due to the lack of explicability and interpretability in ANNs, the analysis of input factors cannot be explained directly by model composition. In this research, we apply a combination of a boosting strategy and input sensitivity analysis in an improved Garson's algorithm to perform feature selection that can adjust its selection criteria through each iteration on an ANN model. This proposed technique is then compared with other traditional feature selection techniques using synthetic data and real-world house valuation data. The results show that our model can maintain the sensitivity coefficient for every informative feature. The technique of this study provides a set of features that influences the house price and implies the character of each specific area. It also contributes to statistical improvement in most valuation cases, with a lower error than that obtained without a selected feature dataset, and it placed in the top 24% of the public leaderboard in the Zillow Prize competition.