دانلود رایگان مقاله انگلیسی کمی سازی عدم قطعیت مبتنی بر شبکه عصبی - IEEE 2018

ABSTRACT

Uncertainty quantification plays a critical role in the process of decision making and optimization in many fields of science and engineering. The field has gained an overwhelming attention among researchers in recent years resulting in an arsenal of different methods. Probabilistic forecasting and in particular prediction intervals (PIs) are one of the techniques most widely used in the literature for uncertainty quantification. Researchers have reported studies of uncertainty quantification in critical applications such as medical diagnostics, bioinformatics, renewable energies, and power grids. The purpose of this survey paper is to comprehensively study neural network-based methods for construction of prediction intervals. It will cover how PIs are constructed, optimized, and applied for decision-making in presence of uncertainties. Also, different criteria for unbiased PI evaluation are investigated. The paper also provides some guidelines for further research in the field of neural network-based uncertainty quantification.

CONCLUSION

The prediction is one of the oldest tasks in this world. The point prediction is the most widely used and understandable form of prediction. However, the point prediction does not convey any information about the uncertainty. Therefore, probabilistic forecasting is becoming popular to predict uncertainties in emerging engineering problems with high uncertainty. The NN based approaches for the quantification of uncertainty is relatively new. Without a thorough discussion, it is impossible to know advantages and disadvantages of different NN based PI construction techniques. The paper provides detailed research activities on NN-based PI construction concerning motives of each work. Our current survey may help analysts in knowing recently available popular NN based PI construction techniques. They may select an appropriate PI construction technique for their application. The work may also help future searchers in developing novel algorithms for constructing NN-based PIs and investigating new applications.