Efficient Fiber Operation of High Ranked Tensors for Dynamic Dataset

Multidimensional data is available in all respects around us, therefore creating the essentiality of high ranked tensors. We will construct a data structure on primary memory to contain the high order tensors in a shape of two dimensional array. The tensor will be shaped with dynamic dataset i.e. data can be compacted later with a data structure capable of expanding. The focus of this paper is to estimate the efficiency of this Stretching Data Structure (SDS) comparing with Traditional Data Structure (TDS) based on fiber operation of high ranked tensors. Because of efficient indexing in SDS we will get a faster data retrieval on fiber. With the increase in dimension the efficiency rate is expected to get improved. Large and complex computation with high order tensors can require immense time hence this research will assist to conduct those computation in a feasible way.