Complex Model, Sequential immigrants, Big Data Generation, Large Sparse Simulation And High-performance Computing Platform for 4iR Applications
A.P Dr Norma Binti Alias
Center for Sustainable Nanomaterials,
Ibnu Sina Institute for Scientific and Industrial Research,
Universiti Teknologi Malaysia
81310, UTM Skudai, Johor , Malaysia
Naturally, the mathematical process starts from proving the existence and uniqueness by the theorem, corollary, lemma, proposition, dealing with the simple and non-complex model. Proving the existence and uniqueness solution is guaranteed by theorem, governing the infinite amount of solutions and limited to the implementation of a small-scale simulation on a single desktop CPU or laptop. Thus, the transition from a small scale to a large sparse simulation from numerical perspective is required for the physical cyber system development for fourth industrial revolution (4iR) applications.
In numerical perspective, the complex model, discretization strategy, sequential immigrants and parallel algorithm can be generated large amounts of data at a very fine granular level in terms of space, time and some dependent parameters. However, there is a lack of researchers on the classification strategy has been done to characterize the extension version of model and methodology for generating big scale from small-scale data and to classify the type of sequential immigrants, parallel algorithms and its simulation to be implemented on the specific taxonomy in high-performance computing (HPC) platform.
The methodology for the classification involving the specific extension in terms of the complex model, derivative and discretization, dimension of space and time, behavior of initial and boundary condition, data generation, extraction, numerical method and image with the high-resolution feature are addressed. The main objective of this proposal is to classify and strategies the grid generation from a small scale concept from numerical perspectives. The outcome of this research is the framework for classification and beneficial for the big data provider, algorithm provider and system analyzer to classify and recommend a specific strategy for generating, handling and analyzing the big data.
Some classification strategies will be investigate integrated with the specific classifier tools. Numerical analysis of sequential algorithm and parallel performance evaluation of parallel algorithm are the indicators for performance investigation of the classification strategy. This research will benefit to make accurate decisions, predictions and trending practice on how to obtain the approximation solution for science and engineering applications. In addition, strategies for classification of generating a big data, identifying the root causes of failures, issues in real time and fully understanding the potential of big data-driven for 4iR applications.