Matrix compression and traversal method and electronic device thereof

A method of traversing a compressed matrix, the method comprising:

• obtaining a data array corresponding to a compressed matrix from at least one memory;
• obtaining, from the at least one memory, a pointer and an offset corresponding to the compressed matrix, wherein the pointer comprises at least one of a row pointer and a column pointer, and wherein the offset comprises at least one of a row offset and a column offset;
• obtaining a data index corresponding to a non-zero data element included in the data array based on an offset corresponding to a traversal direction, a pointer corresponding to the traversal direction, and a pointer corresponding to a direction different from the traversal direction;
• obtaining a value corresponding to the non-zero data element based on the data index;
• and
• reconstructing at least a portion of a decompressed matrix corresponding to the compressed matrix based on the value.

An electronic device configured to perform at least one of matrix compression and matrix traversal, the electronic device comprising:

• one or more processors;
• and
• memory configured to store instructions which, when executed by the one or more processors, individually or collectively, cause the electronic device to:
• obtain an input matrix comprising a plurality of data elements;
• identify a plurality of non-zero data elements in the input matrix and generate a data array in which the plurality of non-zero data elements are arranged sequentially;
• generate a pointer corresponding to each row from among a plurality of rows included in the input matrix, or corresponding to each column from among a plurality of columns included in the input matrix,
• designate, to the pointer, a position in the input matrix or a number of non-zero data elements included in a corresponding row from among the plurality of rows, or included in a corresponding column from among the plurality of columns, wherein the pointer comprises at least one of a row pointer and a column pointer;
• and
• generate an offset that indicates a relative interval between the plurality of non-zero data elements, and designate, to the offset, a relative index for traversing the input matrix in a row direction or a column direction, wherein the offset comprises at least one of a row offset and a column offset,
• generate a compressed matrix corresponding to the input matrix based on the data array, the pointer, and the offset;
• and
• store the compressed matrix in at least one memory.