Sensor-driven Learning of Time-Dependent Parameters for Prescriptive Analytics

Abstract

Big data analytics is rapidly emerging as a key Internet of Things (IoT) initiative aiming at providing meaningful insights and supporting optimal decision making under time constraints. In this direction, prescriptive analytics has just started to emerge. Prescriptive analytics moves beyond descriptive and predictive analytics aiming at providing adaptive, automated, constrained, time-dependent and optimal decisions. The use of time-dependent parameters in prescriptive analytics models provide a more reliable and realistic representation of the complex and dynamic environment and the associated decision making process; however, their estimation poses significant challenges due to the uncertainty derived from inaccurate user input, noisy data, and non-stationarity of real-world data streams. Since feedback and learning mechanisms for tracking the prescriptive analytics are crucial enablers for self-configuration and self-optimization, this paper proposes an approach for sensor-driven learning of time-dependent parameters for prescriptive analytics models deployed in streaming computational environments. The proposed approach was validated in an Industry 4.0 use case, while it was further evaluated through extensive simulation experiments. The proposed approach overcomes challenges related to uncertainty derived from user's input, non-stationary data and sensor noise and provides estimates of time-dependent parameters that lead to more reliable prescriptions. INDEX TERMS Big data, machine learning, data analytics, Internet of Things, non-stationary time-series.

An estimation of effects of memory and learning experience on the eoq model with price dependent demand

“…The theories of learning were implemented on the study of inventory control problems in different ways.In this context, Kazemiet al [56] incorporated the sense of human learning to study an economic order quantity (EOQ) model with back order under fuzzy uncertainty. Also,Shekarianet al [57] considered the fuzzy uncertainty and learning to discuss an EOQ model for the items of imperfect quality.Bousdekiset al [58] developed sensor driven learning approach for the perspective analysis of time dependent parameters that leads to self-optimization through learning and feedback.An alternative intuition for the fuzzy learning-based decision making was established through the introduction of the triangular dense fuzzy set (TDFS) byDe and Beg [59]. The concept of experience-based learning was further enriched with the key and lock facility in terms of the triangular lock fuzzy dense set (TLFDS)by De [60].…”

Discussing Relations Between Dynamic Business Environments and Big Data Analytics

“…These data streams, which vary in volume and velocity and require a continuous processing, are, for instance, of importance in stock markets [56] and the tactical domain [57]. Especially in these fields of application, the focus is less on solely descriptive or predictive analyses, but rather the acquisition of prescriptive knowledge [58]. This means an answer to the question "What should I do?"…”

“…Two consequences for big data analytics applying companies arise from the previous observations on dynamic environments. First, to be able to provide a reasonably priced as well as highly performant solution, businesses need to take advantage of the possibilities of the Internet of Things domain [58] by including the data generating devices into the processing. Thus, dynamics in volume and velocity can be immediately handled by proportionally increasing processing capabilities.…”

“…One way to address this problem according to researchers is the enactment of artificial intelligence [53], [58], [62]. Generally, the objective is to develop a software solution which, after an initial training phase, is capable of self-configuration and fine tuning without human intervention [53].…”

“…At some point, human workforce would not be capable of integrating new tailored algorithmic solutions as quickly as new requirements arise. Since the self-configuration of prescriptive analytics models is a largely unexplored field of research [58], businesses need to strive further to develop and test such technologies to get an edge over their competitors.…”