Search

Utilizing a turnkey approach to deliver a construction project entails significant risks from the contractor's perspective. Essentially, the owner awaits project completion without commitments regarding additional expenditures incurred by the contractor during the project's duration. This paper specifically focuses on estimating and analyzing the contingency value for residential turnkey projects in Saudi Arabia. The contingency value across the project's life cycle is estimated using six Artificial Neural Network (ANN) models, which are compared to identify the best-trained network according to project complexity, contingency factor, and contingency impact during the project phases. The output layer provides the contingency factor percentages for each project phase. A 13-story reinforced concrete (RC) residential building established in one of Saudi Arabia's cities was selected to implement the developed methodology. The contingency estimation, performed using @Risk 7.5 and NeuralTools 7.5, was determined to be 11.34% and was distributed across the five phases of the project's life cycle: 0.30% for predesign, 0.99% for design, 2.61% for preconstruction, 6.33% for construction, and 1.12% for postconstruction. Furthermore, it was found that the estimated contingency varies based on project complexity, which is 7.20% for low complexity, 8.16% for medium complexity, 9.41% for complicated, and 11.34% for very complicated projects. Historical data and peer review approaches are employed to validate the results, both of which are endorsed by professionals in this field. This paper highlights two main contributions: Firstly, it significantly enhances risk management by facilitating a comprehensive understanding and systematic analysis of risks, thus improving the contractors' ability to mitigate potential negative impacts on projects. Secondly, it supports more informed decision-making through the use of advanced techniques to estimate and analyze contingency values. These contributions are critical for contractors engaged in Saudi construction projects, particularly those involving residential buildings. [ABSTRACT FROM AUTHOR]

The acceleration of house technology via the use of mobile phones has made it easier to control houses, where occupants (especially older people) spend most of their time. The climate of Saudi Arabia, especially in the northern area, is too hot during summer and cold during winter. Control of the indoor environment in a smart home is a preferable choice that can reduce power consumption to operate heating, ventilation, and air-conditioning. Machine learning algorithms have been used to predict physical variables of indoor environment, such as temperature and humidity. The model can be trained, learn, and make predictions using historical data. Machine learning techniques can automate temperature monitoring and control. This paper proposes an algorithm that combines Artificial Neural Networks (ANNs) and GridSearchCV to predict physical variables in indoor environments in Saudi Arabia. GridSearchCV was utilized to tune the parameters of the machine learning algorithm. The assessment of the proposed algorithm involved its performance comparison to state-of-the-art machine learning algorithms. A real-world dataset was generated to estimate the performance of the considered algorithms. The room data were collected every 5 min for 31 days during July 2022. The dataset contains 6 columns and 8,910 records from 6 sensors (timestamps, light, temperature, humidity, pressure, and altitude). Random Forest (RF), Decision Tree (DT), and ANN methods were compared with the proposed algorithm. The RF had the highest R2 value of 0.84 and the lowest Mean Square Error (MSE) of 0.43. The DT achieved an R2 score of 0.78, while the ANN achieved R2 score of 0.61, MSE of 1.04, and Mean Absolute Error (MAE) of 0.75. The proposed algorithm achieved an R2 of 0.69, MSE of 0.87, and MAE of 0.67. [ABSTRACT FROM AUTHOR]

Efficient management of hospital resources is essential for providing high-quality healthcare while ensuring sustainability. Length of stay (LOS), measuring the duration from admission to discharge, directly impacts patient outcomes and resource utilization. Accurate LOS prediction offers numerous benefits, including reducing re-admissions, ensuring appropriate staffing, and facilitating informed discharge planning. While conventional methods rely on statistical models and clinical expertise, recent advances in machine learning (ML) present promising avenues for enhancing LOS prediction. This research focuses on developing an ML-based LOS prediction model trained on a comprehensive real-world dataset and discussing the important factors towards practical deployment of trained ML models in clinical settings. This research involves the development of a comprehensive adult cardiac patient dataset (SaudiCardioStay (SCS)) from the King Faisal Specialist Hospital & Research Centre (KFSH&RC) hospital in Saudi Arabia, comprising 4930 patient encounters for 3611 unique patients collected from 2019 to 2022 (excluding 2020). A diverse range of classical ML models (i.e., Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LGBM), artificial neural networks (ANNs), Average Voting Regression (AvgVotReg)) are implemented for the SCS dataset to explore the potential of existing ML models in LOS prediction. In addition, this study introduces a novel approach for LOS prediction by incorporating a dedicated LOS classifier within a sophisticated ensemble methodology (i.e., Two-Level Sequential Cascade Generalization (2LSCG), Three-Level Sequential Cascade Generalization (3LSCG), Parallel Cascade Generalization (PCG)), aiming to enhance prediction accuracy and capture nuanced patterns in healthcare data. The experimental results indicate the best mean absolute error (MAE) of 0.1700 for the 3LSCG model. Relatively comparable performance was observed for the AvgVotReg model, with a MAE of 0.1703. In the end, a detailed analysis of the practical implications, limitations, and recommendations concerning the deployment of ML approaches in actual clinical settings is presented. [ABSTRACT FROM AUTHOR]