Prediction Of Future Consumption Research Articles

A reliable and ensemble forecasting model for slow-moving and repairable spare parts: Data mining approach

Reliable prediction of spare parts is one of the most prevailing challenges for the manufacturing industry and equipment owners, which acts as a double-edged sword. If the objective function focuses on the lack of spare parts, it will cause capital sleep, and otherwise, it would make spare parts shortage. Therefore, the accuracy and reliability of this prediction have always been consequential. Howbeit, the non-uniformity of the spare parts in the real world is remarkable. Some of these parts have intermittent uses (fast-moving type), and some have less and irregular consumption (slow-moving type). Due to this uneven structure of slow-moving spare parts, the application of classical methods has low reliability and performance in forecasting. Therefore, this study concentrated on predicting slow-moving spare parts by developing a reliable and ensemble data mining approach with considering the managerial characteristics (e.g., repair ability and the irregularity of maintenance such as MTBF and MTTR…) along with their history of consumption (the amount of each order and the time interval between two charges). In the proposed approach, the amount of spare parts and the time of the following order is predicted based on an ensemble data mining model from the prediction of future consumption and the probability of scraping in the repair process. This simultaneous attention to data types (consumption and repairs) for slow-moving parts improves data adequacy. Whereas the focus of the literature has been more on improving the data mining models’ performance than on improving data adequacy; by emphasizing both issues at the same time, we can get more reliable results in predicting parts requests with more fluctuations in consumption. A clustering management model for classifying spare parts has been presented during the data preparation process to implement this model. And 51,335 slow-moving spare parts of a Steel Company have been organized and predicted. The results of this study reveal that the application of the proposed approach and ensemble model significantly increases the reliability of spare parts inventory (accuracy by up to 80% and newly defined reliability (RI) by up to 10%).

The prediction of residential building consumption using profiling and time encoding

The prediction of future consumption is vital for building management in general and green buildings in particular. A certified green building should meet an evaluation criterion of resource efficiency throughout its life cycle. This paper presents a hybrid model that combines a profile averaging model and a machine learning model, namely random forest regressor method, for improved far horizon forecasting (10% improvement when the prediction horizon is 1 week). The proposed hybrid model is modular, explainable, and does not entail a large dataset. The feature analysis highlighted the importance of explicit time encoding by feeding the time as an input feature to the forecasting model (importance factor of 35%).

Patient-specific COVID-19 resource utilization prediction using fusion AI model

The strain on healthcare resources brought forth by the recent COVID-19 pandemic has highlighted the need for efficient resource planning and allocation through the prediction of future consumption. Machine learning can predict resource utilization such as the need for hospitalization based on past medical data stored in electronic medical records (EMR). We conducted this study on 3194 patients (46% male with mean age 56.7 (±16.8), 56% African American, 7% Hispanic) flagged as COVID-19 positive cases in 12 centers under Emory Healthcare network from February 2020 to September 2020, to assess whether a COVID-19 positive patient’s need for hospitalization can be predicted at the time of RT-PCR test using the EMR data prior to the test. Five main modalities of EMR, i.e., demographics, medication, past medical procedures, comorbidities, and laboratory results, were used as features for predictive modeling, both individually and fused together using late, middle, and early fusion. Models were evaluated in terms of precision, recall, F1-score (within 95% confidence interval). The early fusion model is the most effective predictor with 84% overall F1-score [CI 82.1–86.1]. The predictive performance of the model drops by 6 % when using recent clinical data while omitting the long-term medical history. Feature importance analysis indicates that history of cardiovascular disease, emergency room visits in the past year prior to testing, and demographic factors are predictive of the disease trajectory. We conclude that fusion modeling using medical history and current treatment data can forecast the need for hospitalization for patients infected with COVID-19 at the time of the RT-PCR test.

A theoretical approach to the Suess effect

From a theoretical study of the production and distribution of fossil-fuel carbon dioxide released since the beginning of the Industrial Revolution calculations have been made of the resultant decreases in atmospheric carbon-14 specific activities (Suess effect). The calculations are based on recent advances in the assessment of parameters which control carbon circulation and a re-evaluation of the combustion rates of fossil fuels. Results show that the reduction in carbon-14 specific activities amounted to -0.5, -3.2 and -5.9 % in A. D. 1890, 1950 and 1969 respectively. Analyses of biospheric materials of known age show good agreement between the predicted and observed atmospheric carbon -14 concentrations although the possibility exists of a perturbation of natural origin. The incorporation of significant amounts of fossil fuel carbon into 1890 wood indicates the possibility of error in radiocarbon analyses based on the conventional modern reference material. The study also enables corrections to be made for the Suess effect to observations of carbon -14 activities of samples grown during the past century. Predictions of future consumption of fossil fuels have permitted evaluation of the approximate magnitude of the future Suess effect. The results imply that the effect will be of increasing importance, reaching about -23% by 2000 and -50 % by 2025. The Suess effect, which is in competition with the nuclear bomb effect, may therefore reduce atmospheric carbon -14 concentration to the natural level again by 1990. Future increase in atmospheric carbon dioxide levels could have significant climatological consequences, but the magnitude of these changes remains speculative.