Search results for: MAPE

Authors: Kudzanayi Chiteka, Wellington Makondo

Abstract:

The whole world is on the drive to go green owing to the negative effects of burning fossil fuels. Therefore, there is immediate need to identify and utilise alternative renewable energy sources. Among these energy sources solar energy is one of the most dominant in Zimbabwe. Solar power plants used to generate electricity are entirely dependent on solar radiation. For planning purposes, solar radiation values should be known in advance to make necessary arrangements to minimise the negative effects of the absence of solar radiation due to cloud cover and other naturally occurring phenomena. This research focused on the prediction of Global Horizontal Irradiation values for the sixth day given values for the past five days. Artificial intelligence techniques were used in this research. Three models were developed based on Support Vector Machines, Radial Basis Function, and Feed Forward Back-Propagation Artificial neural network. Results revealed that Support Vector Machines gives the best results compared to the other two with a mean absolute percentage error (MAPE) of 2%, Mean Absolute Error (MAE) of 0.05kWh/m²/day root mean square (RMS) error of 0.15kWh/m²/day and a coefficient of determination of 0.990. The other predictive models had prediction accuracies of MAPEs of 4.5% and 6% respectively for Radial Basis Function and Feed Forward Back-propagation Artificial neural network. These two models also had coefficients of determination of 0.975 and 0.970 respectively. It was found that prediction of GHI values for the future days is possible using artificial intelligence-based predictive models.

Keywords: solar energy, global horizontal irradiation, artificial intelligence, predictive models

Procedia PDF Downloads 249

Authors: Martins Y. Otache, John J. Musa, Abayomi I. Kuti, Mustapha Mohammed

Abstract:

The performance of an artificial neural network (ANN) is contingent on a host of factors, for instance, the network optimisation scheme. In view of this, the study examined the general implications of the ANN training optimisation algorithm on its forecast performance. To this end, the Bayesian regularisation (Br), Levenberg-Marquardt (LM), and the adaptive learning gradient descent: GDM (with momentum) algorithms were employed under different ANN structural configurations: (1) single-hidden layer, and (2) double-hidden layer feedforward back propagation network. Results obtained revealed generally that the gradient descent with momentum (GDM) optimisation algorithm, with its adaptive learning capability, used a relatively shorter time in both training and validation phases as compared to the Levenberg- Marquardt (LM) and Bayesian Regularisation (Br) algorithms though learning may not be consummated; i.e., in all instances considering also the prediction of extreme flow conditions for 1-day and 5-day ahead, respectively especially using the ANN model. In specific statistical terms on the average, model performance efficiency using the coefficient of efficiency (CE) statistic were Br: 98%, 94%; LM: 98 %, 95 %, and GDM: 96 %, 96% respectively for training and validation phases. However, on the basis of relative error distribution statistics (MAE, MAPE, and MSRE), GDM performed better than the others overall. Based on the findings, it is imperative to state that the adoption of ANN for real-time forecasting should employ training algorithms that do not have computational overhead like the case of LM that requires the computation of the Hessian matrix, protracted time, and sensitivity to initial conditions; to this end, Br and other forms of the gradient descent with momentum should be adopted considering overall time expenditure and quality of the forecast as well as mitigation of network overfitting. On the whole, it is recommended that evaluation should consider implications of (i) data quality and quantity and (ii) transfer functions on the overall network forecast performance.

Keywords: streamflow, neural network, optimisation, algorithm

Procedia PDF Downloads 116

Authors: Shalaka Kadam, Dinesh Apte, Sagar Mainkar

Abstract:

The life cycle of products is becoming shorter and shorter due to increased competition in market, shorter product development time and increased product diversity. Short life cycles are normal in retail industry, style business, entertainment media, and telecom and semiconductor industry. The subject of accurate forecasting for demand of short lifecycle products is of special enthusiasm for many researchers and organizations. Due to short life cycle of products the amount of historical data that is available for forecasting is very minimal or even absent when new or modified products are launched in market. The companies dealing with such products want to increase the accuracy in demand forecasting so that they can utilize the full potential of the market at the same time do not oversupply. This provides the challenge to develop a forecasting model that can forecast accurately while handling large variations in data and consider the complex relationships between various parameters of data. Many statistical models have been proposed in literature for forecasting time series data. Traditional time series forecasting models do not work well for short life cycles due to lack of historical data. Also artificial neural networks (ANN) models are very time consuming to perform forecasting. We have studied the existing models that are used for forecasting and their limitations. This work proposes an effective and powerful forecasting approach for short life cycle time series forecasting. We have proposed an approach which takes into consideration different scenarios related to data availability for short lifecycle products. We then suggest a methodology which combines statistical analysis with structured judgement. Also the defined approach can be applied across domains. We then describe the method of creating a profile from analogous products. This profile can then be used for forecasting products with historical data of analogous products. We have designed an application which combines data, analytics and domain knowledge using point-and-click technology. The forecasting results generated are compared using MAPE, MSE and RMSE error scores. Conclusion: Based on the results it is observed that no one approach is sufficient for short life-cycle forecasting and we need to combine two or more approaches for achieving the desired accuracy.

Keywords: forecast, short life cycle product, structured judgement, time series

Procedia PDF Downloads 323

Authors: Haya Salah, Srinivas Sharan

Abstract:

Healthcare facilities use appointment systems to schedule their appointments and to manage access to their medical services. With the growing demand for outpatient care, it is now imperative to manage physician's time effectively. However, high variation in consultation duration affects the clinical scheduler's ability to estimate the appointment duration and allocate provider time appropriately. Underestimating consultation times can lead to physician's burnout, misdiagnosis, and patient dissatisfaction. On the other hand, appointment durations that are longer than required lead to doctor idle time and fewer patient visits. Therefore, a good estimation of consultation duration has the potential to improve timely access to care, resource utilization, quality of care, and patient satisfaction. Although the literature on factors influencing consultation length abound, little work has done to predict it using based data-driven approaches. Therefore, this study aims to predict consultation duration using supervised machine learning algorithms (ML), which predicts an outcome variable (e.g., consultation) based on potential features that influence the outcome. In particular, ML algorithms learn from a historical dataset without explicitly being programmed and uncover the relationship between the features and outcome variable. A subset of the data used in this study has been obtained from the electronic medical records (EMR) of four different outpatient clinics located in central Pennsylvania, USA. Also, publicly available information on doctor's characteristics such as gender and experience has been extracted from online sources. This research develops three popular ML algorithms (deep learning, random forest, gradient boosting machine) to predict the treatment time required for a patient and conducts a comparative analysis of these algorithms with respect to predictive performance. The findings of this study indicate that ML algorithms have the potential to predict the provider service time with superior accuracy. While the current approach of experience-based appointment duration estimation adopted by the clinic resulted in a mean absolute percentage error of 25.8%, the Deep learning algorithm developed in this study yielded the best performance with a MAPE of 12.24%, followed by gradient boosting machine (13.26%) and random forests (14.71%). Besides, this research also identified the critical variables affecting consultation duration to be patient type (new vs. established), doctor's experience, zip code, appointment day, and doctor's specialty. Moreover, several practical insights are obtained based on the comparative analysis of the ML algorithms. The machine learning approach presented in this study can serve as a decision support tool and could be integrated into the appointment system for effectively managing patient scheduling.

Keywords: clinical decision support system, machine learning algorithms, patient scheduling, prediction models, provider service time

Procedia PDF Downloads 87

Authors: S. Sabatino, V. Calderaro, V. Galdi, G. Graber, L. Ippolito

Abstract:

Climate change is a rapidly growing global threat caused mainly by increased emissions of carbon dioxide (CO₂) into the atmosphere. These emissions come from multiple sources, including industry, power generation, and the transport sector. The need to tackle climate change and reduce CO₂ emissions is indisputable. A crucial solution to achieving decarbonization in the transport sector is the adoption of electric vehicles (EVs). These vehicles use lithium (Li-Ion) batteries as an energy source, making them extremely efficient and with low direct emissions. However, Li-Ion batteries are not without problems, including the risk of overheating and performance degradation. To ensure its safety and longevity, it is essential to use a battery management system (BMS). The BMS constantly monitors battery status, adjusts temperature and cell balance, ensuring optimal performance and preventing dangerous situations. From the monitoring carried out, it is also able to optimally manage the battery to increase its life. Among the parameters monitored by the BMS, the main ones are State of Charge (SoC), State of Health (SoH), and State of Power (SoP). The evaluation of these parameters can be carried out in two ways: offline, using benchtop batteries tested in the laboratory, or online, using batteries installed in moving vehicles. Online estimation is the preferred approach, as it relies on capturing real-time data from batteries while operating in real-life situations, such as in everyday EV use. Actual battery usage conditions are highly variable. Moving vehicles are exposed to a wide range of factors, including temperature variations, different driving styles, and complex charge/discharge cycles. This variability is difficult to replicate in a controlled laboratory environment and can greatly affect performance and battery life. Online estimation captures this variety of conditions, providing a more accurate assessment of battery behavior in real-world situations. In this article, a hybrid approach based on a neural network and a statistical method for real-time estimation of SoC, SoH, and SoP parameters of interest is proposed. These parameters are estimated from the analysis of a one-day driving profile of an electric vehicle, assumed to be divided into the following four phases: (i) Partial discharge (SoC 100% - SoC 50%), (ii) Partial discharge (SoC 50% - SoC 80%), (iii) Deep Discharge (SoC 80% - SoC 30%) (iv) Full charge (SoC 30% - SoC 100%). The neural network predicts the values of ohmic resistance and incremental capacity, while the statistical method is used to estimate the parameters of interest. This reduces the complexity of the model and improves its prediction accuracy. The effectiveness of the proposed model is evaluated by analyzing its performance in terms of square mean error (RMSE) and percentage error (MAPE) and comparing it with the reference method found in the literature.

Keywords: electric vehicle, Li-Ion battery, BMS, state-of-charge, state-of-health, state-of-power, artificial neural networks

Procedia PDF Downloads 39

Authors: Elizaveta S. Prokofyeva, Svetlana V. Maltseva, Roman D. Zaitsev

Abstract:

Rapidly evolving modern data analysis technologies in healthcare play a large role in understanding the operation of the system and its characteristics. Nowadays, one of the key tasks in urban healthcare is to optimize the resource allocation. Thus, the application of data analysis in medical institutions to solve optimization problems determines the significance of this study. The purpose of this research was to establish the dependence between the indicators of the effectiveness of the medical institution and its resources. Hospital discharges by diagnosis; hospital days of in-patients and in-patient average length of stay were selected as the performance indicators and the demand of the medical facility. The hospital beds by type of care, medical technology (magnetic resonance tomography, gamma cameras, angiographic complexes and lithotripters) and physicians characterized the resource provision of medical institutions for the developed models. The data source for the research was an open database of the statistical service Eurostat. The choice of the source is due to the fact that the databases contain complete and open information necessary for research tasks in the field of public health. In addition, the statistical database has a user-friendly interface that allows you to quickly build analytical reports. The study provides information on 28 European for the period from 2007 to 2016. For all countries included in the study, with the most accurate and complete data for the period under review, predictive models were developed based on historical panel data. An attempt to improve the quality and the interpretation of the models was made by cluster analysis of the investigated set of countries. The main idea was to assess the similarity of the joint behavior of the variables throughout the time period under consideration to identify groups of similar countries and to construct the separate regression models for them. Therefore, the original time series were used as the objects of clustering. The hierarchical agglomerate algorithm k-medoids was used. The sampled objects were used as the centers of the clusters obtained, since determining the centroid when working with time series involves additional difficulties. The number of clusters used the silhouette coefficient. After the cluster analysis it was possible to significantly improve the predictive power of the models: for example, in the one of the clusters, MAPE error was only 0,82%, which makes it possible to conclude that this forecast is highly reliable in the short term. The obtained predicted values of the developed models have a relatively low level of error and can be used to make decisions on the resource provision of the hospital by medical personnel. The research displays the strong dependencies between the demand for the medical services and the modern medical equipment variable, which highlights the importance of the technological component for the successful development of the medical facility. Currently, data analysis has a huge potential, which allows to significantly improving health services. Medical institutions that are the first to introduce these technologies will certainly have a competitive advantage.

Keywords: data analysis, demand modeling, healthcare, medical facilities

Procedia PDF Downloads 113

Authors: Selin Guney, Andres Riquelme

Abstract:

Among the main purposes of optimal and efficient forecasts of agricultural commodity prices is to guide the firms to advance the economic decision making process such as planning business operations and marketing decisions. Governments are also the beneficiaries and suppliers of agricultural price forecasts. They use this information to establish a proper agricultural policy, and hence, the forecasts affect social welfare and systematic errors in forecasts could lead to a misallocation of scarce resources. Various empirical approaches have been applied to forecast commodity prices that have used different methodologies. Most commonly-used approaches to forecast commodity sectors depend on classical time series models that assume values of the response variables are precise which is quite often not true in reality. Recently, this literature has mostly evolved to a consideration of fuzzy time series models that provide more flexibility in terms of the classical time series models assumptions such as stationarity, and large sample size requirement. Besides, fuzzy modeling approach allows decision making with estimated values under incomplete information or uncertainty. A number of fuzzy time series models have been developed and implemented over the last decades; however, most of them are not appropriate for forecasting repeated and nonconsecutive transitions in the data. The modeling scheme used in this paper eliminates this problem by introducing Markov modeling approach that takes into account both the repeated and nonconsecutive transitions. Also, the determination of length of interval is crucial in terms of the accuracy of forecasts. The problem of determining the length of interval arbitrarily is overcome and a methodology to determine the proper length of interval based on the distribution or mean of the first differences of series to improve forecast accuracy is proposed. The specific purpose of this paper is to propose and investigate the potential of a new forecasting model that integrates methodologies for determining the proper length of interval based on the distribution or mean of the first differences of series and Fuzzy Time Series- Markov Chain model. Moreover, the accuracy of the forecasting performance of proposed integrated model is compared to different univariate time series models and the superiority of proposed method over competing methods in respect of modelling and forecasting on the basis of forecast evaluation criteria is demonstrated. The application is to daily corn and soybean prices observed at three commercially important North Carolina markets; Candor, Cofield and Roaring River for corn and Fayetteville, Cofield and Greenville City for soybeans respectively. One main conclusion from this paper is that using fuzzy logic improves the forecast performance and accuracy; the effectiveness and potential benefits of the proposed model is confirmed with small selection criteria value such MAPE. The paper concludes with a discussion of the implications of integrating fuzzy logic and nonarbitrary determination of length of interval for the reliability and accuracy of price forecasts. The empirical results represent a significant contribution to our understanding of the applicability of fuzzy modeling in commodity price forecasts.

Keywords: commodity, forecast, fuzzy, Markov

Procedia PDF Downloads 196