Search results

Accurate and timely cost prediction is critical to the success of construction projects which is still facing challenges especially at the early stage. In the context of rapid development of machine learning technology and the massive cost data from historical projects, this paper aims to propose a novel cost prediction model based on historical data with improved performance when only limited information about the new project is available.

The proposed approach combines regression analysis (RA) and artificial neural network (ANN) to build a novel hybrid cost prediction model with the former as front-end prediction and the latter as back-end correction. Firstly, the main factors influencing the cost of building projects are identified through literature research and subsequently screened by principal component analysis (PCA). Secondly the optimal RA model is determined through multi-model comparison and used for front-end prediction. Finally, ANN is applied to construct the error correction model. The hybrid RA-ANN model was trained and tested with cost data from 128 completed construction projects in China.

The results show that the hybrid cost prediction model has the advantages of both RA and ANN whose prediction accuracy is higher than that of RA and ANN only with the information such as total floor area, height and number of floors.

(1) The most critical influencing factors of the buildings’ cost are found out by means of PCA on the historical data. (2) A novel hybrid RA-ANN model is proposed which proved to have the advantages of both RA and ANN with higher accuracy. (3) The comparison among different models has been carried out which is helpful to future model selection.

In past decades, artificial neural network (ANN) models have revolutionised various stock market operations due to their superior ability to deal with nonlinear data and garnered considerable attention from researchers worldwide. The present study aims to synthesize the research field concerning ANN applications in the stock market to a) systematically map the research trends, key contributors, scientific collaborations, and knowledge structure, and b) uncover the challenges and future research areas in the field.

To provide a comprehensive appraisal of the extant literature, the study adopted the mixed approach of quantitative (bibliometric analysis) and qualitative (intensive review of influential articles) assessment to analyse 1,483 articles published in the Scopus and Web of Science indexed journals during 1992–2022. The bibliographic data was processed and analysed using VOSviewer and R software.

The results revealed the proliferation of articles since 2018, with China as the dominant country, Wang J as the most prolific author, “Expert Systems with Applications” as the leading journal, “computer science” as the dominant subject area, and “stock price forecasting” as the predominantly explored research theme in the field. Furthermore, “portfolio optimization”, “sentiment analysis”, “algorithmic trading”, and “crisis prediction” are found as recently emerged research areas.

To the best of the authors’ knowledge, the current study is a novel attempt that holistically assesses the existing literature on ANN applications throughout the entire domain of stock market. The main contribution of the current study lies in discussing the challenges along with the viable methodological solutions and providing application area-wise knowledge gaps for future studies.

The prediction of stock market (SM) indices is a fascinating task. An in-depth analysis in this field can provide valuable information to investors, traders and policy makers in attractive SMs. This article aims to apply a correlation feature selection model to identify important technical indicators (TIs), which are combined with multiple deep learning (DL) algorithms for forecasting SM indices.

The methodology involves using a correlation feature selection model to select the most relevant features. These features are then used to predict the fluctuations of six markets using various DL algorithms, and the results are compared with predictions made using all features by using a range of performance measures.

The experimental results show that the combination of TIs selected through correlation and Artificial Neural Network (ANN) provides good results in the MADEX market. The combination of selected indicators and Convolutional Neural Network (CNN) in the NASDAQ 100 market outperforms all other combinations of variables and models. In other markets, the combination of all variables with ANN provides the best results.

This article makes several significant contributions, including the use of a correlation feature selection model to select pertinent variables, comparison between multiple DL algorithms (ANN, CNN and Long-Short-Term Memory (LSTM)), combining selected variables with algorithms to improve predictions, evaluation of the suggested model on six datasets (MASI, MADEX, FTSE 100, SP500, NASDAQ 100 and EGX 30) and application of various performance measures (Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error(RMSE), Mean Squared Logarithmic Error (MSLE) and Root Mean Squared Logarithmic Error (RMSLE)).

This paper aims to create and evaluate a model for cryptocurrency adoption by investigating how age, education, and gender impact Behavioural Intention. A hybrid approach that combined partial least squares structural equation modeling (PLS-SEM) and artificial neural network (ANN) was used for the purpose.

This study uses a multi-analytical hybrid approach, combining PLS-SEM and ANN to illustrate the impact of various identified variables on behavioral intention toward using cryptocurrency. Multi-group analysis (MGA) is applied to determine whether different data groups of age, gender and education have significant differences in the parameter estimates that are specific to each group.

The findings indicate that Social Influence (SI) has the greatest impact on Behavioral Intention (BI), which suggests that the viewpoints and recommendations of influential and well-known individuals can serve as a motivating factor to invest in cryptocurrencies. Furthermore, education was found to be a moderating factor in the relationship found between behavioral intention and design.

Prior studies on technology adoption have utilized superficial SEM and ANN methods, whereas a more effective outcome has been suggested by implementing a dual-stage PLS-SEM and ANN approach utilizing a deep neural network architecture. This methodology can enhance the accuracy of nonlinear connections in the model and augment the deep learning capacity.

The research is based on the Unified Theory of Acceptance and Use of Technology (UTAUT2) and expands upon this model by integrating elements of design and trust. This is an important addition, as design can influence individuals' willingness to try new technologies, while trust is a critical factor in determining whether individuals will adopt and use new technology.

Cryptocurrencies are a relatively new phenomenon in India, and their use and adoption have grown significantly in recent years. However, this development has not been without controversy, as the implications of cryptocurrencies for society, the economy and governance remain uncertain. The results reveal that social influence is an important predictor for the adoption of cryptocurrency in India, and this can help financial institutions and regulators in making policy decisions accordingly.

Given the emerging nature of cryptocurrency adoption in India, there is certainly a need for further empirical research in this area. The current study aims to address this research gap and achieve the following objectives: (a) to determine if a dual-stage PLS-SEM and ANN analysis utilizing deep learning techniques can yield more comprehensive research findings than a PLS-SEM approach and (b) to identify variables that can forecast the intention to adopt cryptocurrency.

Cost overrun in infrastructure projects is a constant concern, with a need for a proper solution. The current estimation practice needs improvement to reduce cost overruns. This study aimed to find possible statistical modelling techniques that could be used to develop cost models to produce more reliable cost estimates.

A bibliographic literature review was conducted using a two-stage selection method to compile the relevant publications from Scopus. Then, Visualisation of Similarities (VOS)-Viewer was used to develop the visualisation maps for co-occurrence keyword analysis and yearly trends in research topics.

The study found seven primary techniques used as cost models in construction projects: regression analysis (RA), artificial neural network (ANN), case-based reasoning (CBR), fuzzy logic, Monte-Carlo simulation (MCS), support vector machine (SVM) and reference class forecasting (RCF). RA, ANN and CBR were the most researched techniques. Furthermore, it was observed that the model's performance could be improved by combining two or more techniques into one model.

The research was limited to the findings from the bibliometric literature review.

The findings provided an assessment of statistical techniques that the industry can adopt to improve the traditional estimation practice of infrastructure projects.

This study mapped the research carried out on cost-modelling techniques and analysed the trends. It also reviewed the performance of the models developed for infrastructure projects. The findings could be used to further research to develop more reliable cost models using statistical modelling techniques with better performance.

The objective of this research is to put forward a novel closed-loop circular economy (CE) approach to forecast the sustainability of supply chains (SCs). We provide a practical and real-world CE framework to improve and fill the current knowledge gap in evaluating sustainability of SCs. Besides, we aim to propose a real-life managerial forecasting approach to alert the decision-makers on the future unsustainability of SCs.

It is needed to develop an integrated mathematical model to deal with the complexity of sustainability and CE criteria. To address this necessity, for the first time, network data envelopment analysis (NDEA) is incorporated into the dynamic data envelopment analysis (DEA) and artificial neural network (ANN). In general, methodologically, the paper uses a novel hybrid decision-making approach based on a combination of dynamic and network DEA and ANN models to evaluate sustainability of supply chains using environmental, social, and economic criteria based on real life data and experiences of knowledge-based companies so that the study has a good adaptation with the scope of the journal.

A practical CE evaluation framework is proposed by incorporating recyclable undesirable outputs into the models and developing a new hybrid “dynamic NDEA” and “ANN” model. Using ANN, the sustainability trend of supply chains for future periods is forecasted, and the benchmarks are proposed. We deal with the undesirable recycling outputs, inputs, desirable outputs and carry-overs simultaneously.

We propose a novel hybrid dynamic NDEA and ANN approach for forecasting the sustainability of SCs. To do so, for the first time, we incorporate a practical CE concept into the NDEA. Applying the hybrid framework provides us a new ranking approach based on the sustainability trend of SCs, so that we can forecast unsustainable supply chains and recommend preventive solutions (benchmarks) to avoid future losses. A practicable case study is given to demonstrate the real-life applications of the proposed method.

The continuous increase of energy demands is a critical worldwide matter. Jordan’s household sector accounts for 44% of overall electricity usage annually. This study aims to use artificial neural networks (ANNs) to assess and forecast electricity usage and demands in Jordan’s residential sector.

Four parameters are evaluated throughout the analysis, namely, population (P), income level (IL), electricity unit price (E$) and fuel unit price (F$). Data on electricity usage and independent factors are gathered from government and literature sources from 1985 to 2020. Several networks are analyzed and optimized for the ANN in terms of root mean square error, mean absolute percentage error and coefficient of determination (R²).

The predictions of this model are validated and compared with literature-reported models. The results of this investigation showed that the electricity demand of the Jordanian household sector is mainly driven by the population and the fuel price. Finally, time series analysis approach is incorporated to forecast the electricity demands in Jordan’s residential sector for the next decade.

The paper provides useful recommendations and suggestions for the decision-makers in the country for dynamic planning for future resource policies in the household sector.

Fused Filament Fabrication (FFF) is one of the growing technologies in additive manufacturing, that can be used in a number of applications. In this method, process parameters can be customized and their simultaneous variation has conflicting impacts on various properties of printed parts such as dimensional accuracy (DA) and surface finish. These properties could be improved by optimizing the values of these parameters.

In this paper, four process parameters, namely, print speed, build orientation, raster width, and layer height which are referred to as “input variables” were investigated. The conflicting influence of their simultaneous variations on the DA of printed parts was investigated and predicated. To achieve this goal, a hybrid Genetic Algorithm – Artificial Neural Network (GA-ANN) model, was developed in C#.net, and three geometries, namely, U-shape, cube and cylinder were selected. To investigate the DA of printed parts, samples were printed with a central through hole. Design of Experiments (DoE), specifically the Rotational Central Composite Design method was adopted to establish the number of parts to be printed (30 for each selected geometry) and also the value of each input process parameter. The dimensions of printed parts were accurately measured by a shadowgraph and were used as an input data set for the training phase of the developed ANN to predict the behavior of process parameters. Then the predicted values were used as input to the Desirability Function tool which resulted in a mathematical model that optimizes the input process variables for selected geometries. The mean square error of 0.0528 was achieved, which is indicative of the accuracy of the developed model.

The results showed that print speed is the most dominant input variable compared to others, and by increasing its value, considerable variations resulted in DA. The inaccuracy increased, especially with parts of circular cross section. In addition, if there is no need to print parts in vertical position, the build orientation should be set at 0° to achieve the highest DA. Finally, optimized values of raster width and layer height improved the DA especially when the print speed was set at a high value.

By using ANN, it is possible to investigate the impact of simultaneous variations of FFF machines’ input process parameters on the DA of printed parts. By their optimization, parts of highly accurate dimensions could be printed. These findings will be of significant value to those industries that need to produce parts of high DA on FFF machines.

This study aims at proposing a hybrid model for early cost prediction of a construction project. Early cost prediction for a construction project is the basic approach to procure a project within a predefined budget. However, most of the projects routinely face the impact of cost overruns. Furthermore, conventional and manual cost computing techniques are hectic, time-consuming and error-prone. To deal with such challenges, soft computing techniques such as artificial neural networks (ANNs), fuzzy logic and genetic algorithms are applied in construction management. Each technique has its own constraints not only in terms of efficiency but also in terms of feasibility, practicability, reliability and environmental impacts. However, appropriate combination of the techniques improves the model owing to their inherent nature.

This paper proposes a hybrid model by combining machine learning (ML) techniques with ANN to accurately predict the cost of pile foundations. The parameters contributing toward the cost of pile foundations were collected from five different projects in India. Out of 180 collected data entries, 176 entries were finally used after data cleaning. About 70% of the final data were used for building the model and the remaining 30% were used for validation.

The proposed model is capable of predicting the pile foundation costs with an accuracy of 97.42%.

Although various cost estimation techniques are available, appropriate use and combination of various ML techniques aid in improving the prediction accuracy. The proposed model will be a value addition to cost estimation of pile foundations.

The weld joint mechanical properties of friction stir welding (FSW) are majorly reliant on different input parameters of the FSW machine. The study and optmization of these parameters is uttermost requirement and aim of this study to increase the suitability of FSW in different manufacturing industries. Hence, the input parameters are optimized through different soft computing methods to increase the considered objective in this study.

In this research, ultimate tensile strength (UTS), yield strength (YS) and elongation (EL) of FSW prepared butt joints of AA6061 and AA5083 Aluminium alloys materials are investigated as per American Society for Testing and Materials (ASTM E8-M04) standard. The FSW joints were prepared by changing the three input process parameters. To develop experimental run order design matrix, rotatable central composite design strategy was used. Furthermore, genetic algorithm (GA) in combination (Hybrid) with response surface methodology (RSM), artificial neural network (ANN), i.e. RSM-GA, ANN-GA, is exercised to optimize the considered process parameters.

The maximum value of UTS, YS and EL of test specimens on universal testing machine was measured as 264 MPa, 204 MPa and 14.41%, respectively. The most optimized results (UTS = 269.544 MPa, YS = 211.121 MPa and EL = 17.127%) are obtained with ANN-GA for the considered objectives.

The optimization of input parameters to increase the output objective values using hybrid soft computing techniques is unique in this research paper. The outcomes of this study will help the FSW using manufacturing industries to choose the best optimized parameters set for FSW prepared butt joint with improved mechanical properties.