A Price-Based Grid Resources Pricing Approach for NonStorable Real Assets

Abstract

A grid-computing paradigm delivers the processing power of massively parallel computation to all subscribed users. Current trends, research, and developments in grid computing show that the available grid resources exist as non-storable compute commodities and are distributed geographically – the grid problem. To solve the grid problem, several initiatives have developed frameworks for grid economy and have proposed several algorithms towards an optimized resources scheduling in a grid environment. However, since the grid resources availability depends on the time of usage and are transient, such generic approaches lack the ability to capture the realistic valuation of the resources and fail to guarantee the certainty in their availability measured as Quality of Service (QoS). Uncertainties in grid resource availability do not guarantee a user expected QoS without over committing (e.g., storing the non-storable resources) resources to the users. To guarantee QoS (satisfy a users’ computing needs), we propose a price-based and quality-aware model that captures the realistic value of the grid compute commodities. We use the financial option theory from a real option perspective to value grid resources by treating them as real assets. We discuss a set of results on pricing grid compute cycles. Our results are based on the compute cycle usage obtained from the WestGrid node at the University of Manitoba. We extend and generalize our study to any grid in general but with specific reference to the WestGrid. Keywords: Financial Options, Price Stochasticity, Compute Cycles, Real Options, Quality of Service. Allenotor, D. & Oyemade, D. A. (2022): A Price-Based Grid Resources Pricing Approach for Non-Storable Real AssetsJournal of Advances in Mathematical & Computational Science. Vol. 10, No. 2. Pp 1-18 DOI: dx.doi.org/10.22624/AIMS/MATHS/V10N2P1. Available online at www.isteams.net/mathematics-computationaljournal.

TiSPHiMME: Time Series Profile Hidden Markov Ensemble in Resolving Item Location on Shelf Placement in Basket Analysis

“…Time series clustering can be used to identify products that are commonly purchased across a certain period. Such patterns are otherwise hard to discover using association rule mining, which transactions without temporal consideration (Allenotor et al, 2015;Allenotor & Ojugo, 2017;Mustofa et al, 2023;Ojugo & Otakore, 2020a;Ojugo & Yoro, 2020).…”