Gram-Schmidt Orthogonalization Process in Soft

Inner Product Space

Indubaran Mandal*
Department of Mathematics, Bidhan Chandra College, Asansol-713304

Email: indubaran707@[Link]

___________________________________________________________________________

Abstract:
In this present paper we have studied the Gram-Schmidt Orthogonalization Process, which
allow us to transform a set of linearly independent soft vectors into a set of orthonormal soft
vectors, by analysing the orthogonality of soft vectors in a soft inner product space.
Furthermore, we have discussed some examples and applications of Gram-Schmidt
Orthogonalization Process. The starting point of this work is the basis of soft linear space
and then analysis of the Soft Hilbert spaces and orthonormality in Soft Hilbert spaces are
studied.

AMS Classification: 03E72, 08A72

Keywords: Soft Inner Product Spaces, Orthogonal, Orthonormal, Gram-Schmidt Process.

*Corresponding Author: Indubaran Mandal (indubaran707@[Link])

___________________________________________________________________________

1. Introduction
In the year 1999, Molodtsov [16] initiated the theory of soft sets as a new mathematical tool
for dealing with uncertainties. He has shown several applications of this theory in solving
many practical problems in economics, engineering, social science, medical science, etc.
Research works in soft set theory and its applications in various fields have been progressing
rapidly since Maji et al. ([13],[14]) introduced several operations on soft sets and applied it to
decision making problems. In the line of reduction and addition of parameters of soft sets
some works have been done by Chen [3], Pei and Miao [17] , Kong et al. [12] , Zou and Xiao
[19]. Aktas and Cagman [1] introduced the notion of soft group and discussed various
properties. Jun ([11]) investigated soft BCK/BCI - algebras and its application in ideal theory.
Feng et al. [9] worked on soft semirings, soft ideals and idealistic soft semirings. Ali et al. [2]
and Shabir and Irfan Ali ([2]) studied soft semigroups and soft ideals over a semi group
which characterize generalized fuzzy ideals and fuzzy ideals with thresholds of a semigroup.
The idea of soft topological spaces was first given by M. Shabir, M. Naz [18] and mappings
between soft sets were described by P. Majumdar, S. K. Samanta [15]. Feng et al. [10]
worked on soft sets combined with fuzzy sets and rough sets.

In ([4],[5]) Das and Samanta have introduced a notion of soft real sets, soft real numbers, soft
complex sets, soft complex numbers and some of their basic properties have been
investigated. Some applications of soft real sets and soft real numbers have been presented in
real life problems. In ([6],[7]) Das and Samanta introduced the concept of `soft metric', `soft
linear spaces', `soft norm' on a `soft linear spaces' and studied various properties of `soft
metric spaces' and `soft normed linear spaces' in details. In [8] they have introduced a notion
of soft inner product on soft linear space and studied some of its properties like soft Hilbert
spaces and the concepts of orthogonality and orthonormality in soft Hilbert spaces and their
various properties are studied.

In this paper we have studied the Gram-Schmidt Orthogonalization Process, which allow us
to transform a set of linearly independent soft vectors into a set of orthonormal soft vectors,
by analysing the orthogonality of soft vectors in a soft inner product space. In section 2,
some preliminary results are given. In section 3, some definitions like soft basis , soft
orthogonal set and soft orthonormal sets are given. Then we discussed about the Gram-
Schmidt process in crisp inner product space and main the result i.e. the Gram-Schmidt
process in soft inner product space. Section 4 concludes the paper.

2. Preliminaries
Definition 2.1 (see [16]): Let 𝑈 be an universe set and 𝐸 be a set of parameters. Let 𝑃(𝑈)
denote the power set of 𝑈 and 𝐴 be a non-empty subset of 𝐸. A pair (𝐹, 𝐴) is called a soft set
over 𝑈, where 𝐹 is mapping given by 𝐹: 𝐴 → 𝑃(𝑈). In other words, a soft set over 𝑈 is a
parametrized family of subsets of the universe 𝑈. For 𝜀 ∈ 𝐴, 𝐹(𝜀) may be considered as the
set of 𝜀- approximate elements of the soft set (𝐹, 𝐴) .

Definition 2.2 (see [14]): For two soft sets (𝐹, 𝐴) and (𝐺, 𝐵) over a common universe 𝑈, we
say that (𝐹, 𝐴) is a soft subset of (𝐺, 𝐵) if
(1) 𝐴 ⊆ 𝐵 and
(2) for all 𝑒 ∈ 𝐴, 𝐹(𝑒) ⊆ 𝐺(𝑒)
We write (𝐹, 𝐴) ⊂
̃ (𝐺, 𝐵). (F,A) is said to be a soft super set of (𝐺, 𝐵), if (𝐺, 𝐵) is a soft
subset of (𝐹, 𝐴). We denote it by (𝐹, 𝐴) ⊃
̃ (𝐺, 𝐵).

Definition 2.3 (see [14]): Two soft set (𝐹, 𝐴) and (𝐺, 𝐵) over a common universe 𝑈 are said
to be equal if (𝐹, 𝐴) is a soft subset of (𝐺, 𝐵) and (𝐺, 𝐵) is a soft subset of (𝐹, 𝐴).

Definition 2.4 (see [14]): The complement of a soft set (𝐹, 𝐴) is denoted by (𝐹, 𝐴)𝑐 =
(𝐹 𝑐 , 𝐴), where 𝐹 𝑐 : 𝐴 → 𝑃(𝑈) is a mapping given by 𝐹 𝑐 (𝑒) = 𝑈 − 𝐹(𝑒), for all 𝑒 ∈ 𝐴.

Definition 2.5 (see [14]): A soft set (𝐹, 𝐴) over 𝑈 is said to be a NULL soft set denoted by
Φ if for all 𝑒 ∈ 𝐴, 𝐹(𝑒) = 𝜙 (null set).

Definition 2.6 (see [14]): A soft set (𝐹, 𝐴) over 𝑈 is said to be an absolute soft set denoted by
̆ if for all 𝑒 ∈ 𝐴 , 𝐹(𝑒) = 𝑈.
𝑈

Definition 2.7 (see [4]): Let 𝑈 be a non-empty set and A be a non-empty parameter set. Then
a function 𝜀 ∶ 𝐴 → 𝑈 is said to be a soft element of 𝑈. A soft element 𝜀 of 𝑈 is said to belongs
to a soft set 𝐵 of 𝑈 , which is denoted by 𝜀 ∈
̃ 𝐵, if 𝜀(𝑒) ∈ 𝐵(𝑒), ∀ 𝑒 ∈ 𝐴. Thus for a soft set
𝐵 of 𝑈 with respect to the parameter set 𝐴, we have 𝐵(𝑒) = { 𝜀(𝑒): 𝜀 ∈
̃ 𝐵}, 𝑒 ∈ 𝐴.

It is to be noted that every singleton soft set (a soft set (𝐹, 𝐴) for which 𝐹(𝑒) is a singleton
set, ∀ 𝑒 ∈ 𝐴 ) can be identified with a soft element by simply identifying the singleton set
with the element that it contains 𝑒 ∀ 𝐴.

Definition 2.8 (see [4]): Let ℝ be the set of real numbers and ℬ(ℝ) the collection of all non-
empty bounded subsets of ℝ and 𝐴 taken as a set of parameters. Then a mapping 𝐹: 𝐴 →
ℬ(ℝ) is called soft real set. It is denoted by (𝐹, 𝐴). If specifically (𝐹, 𝐴) is a singleton soft
set, then after identifying (𝐹, 𝐴) with the corresponding soft element, it will be called a soft
real number. The set of all soft real numbers is denoted by ℝ(𝐴).

We use notations 𝑥̃, 𝑦̃, 𝑧̃ to denote soft real numbers whereas 𝑥̅ , 𝑦̅, 𝑧̅ will denote a particular
type of soft real numbers such that 𝑥̅ (𝜆) = 𝑥 for all 𝜆 ∈ 𝐴. For example 0̅ is the soft real
number where 0̅(𝜆) = 0, ∀ 𝜆 ∈ 𝐴.

Definition 2.9 (see [4]): For two soft real numbers 𝑟̃ , 𝑠̃ we define
̃ 𝑠̃ if 𝑟̃ (𝜆) ≤ 𝑠̃ (𝜆) ,for all 𝜆 ∈ 𝐴.
(i) 𝑟̃ ≤
̃ 𝑠̃ if 𝑟̃ (𝜆) ≥ 𝑠̃ (𝜆) ,for all 𝜆 ∈ 𝐴.
(ii) 𝑟̃ ≥
(iii) 𝑟̃ <
̃ 𝑠̃ if 𝑟̃ (𝜆) < 𝑠̃ (𝜆) ,for all 𝜆 ∈ 𝐴.
(iv) 𝑟̃ >
̃ 𝑠̃ if 𝑟̃ (𝜆) > 𝑠̃ (𝜆) ,for all 𝜆 ∈ 𝐴.

Definition 2.10 (see [4]): A soft real number 𝑟̃ is said to be non-negative if 𝑟̃ (𝜆) ≥ 0, ∀ 𝜆 ∈
𝐴. We denote the set of all non-negative soft real numbers by ℝ(𝐴)∗ .

Remark 2.11 (see [7]): Let 𝑋 be a non-empty set. Let 𝑋̆ be absolute set i.e. 𝐹(𝜆) = 𝑋, ∀ 𝜆 ∈
𝐴, where (𝐹, 𝐴) = 𝑋̆ . Let 𝑆(𝑋̆) be the collection of the null soft set Φ and those soft sets
(𝐹, 𝐴) over 𝑋 for which 𝐹(𝜆) ≠ 𝜙, ∀𝜆 ∈ 𝐴. For (𝐹, 𝐴)(≠ Φ) ∈ 𝑆(𝑋̆), the collection of all
soft elements of (𝐹, 𝐴) will be denoted by 𝑆𝐸(𝐹, 𝐴).

Definition 2.12 (see [7]): Let 𝑉 be a vector space over a field 𝐾 and let 𝐴 be a parameter set.
Let (𝐺, 𝐴) is a soft set over 𝑉. Now 𝐺 is said to be a soft vector space or soft linear space of
𝑉 over 𝐾 if 𝐺(𝜆) is a vector space of 𝑉, ∀ 𝜆 ∈ 𝐴.

Definition 2.13 (see [7]): Let 𝐹 be a soft vector space of 𝑉 over 𝐾. Let 𝐺: 𝐴 → 𝑃(𝑉) be a soft
set over 𝑉. Then 𝐺 is said to be a soft vector subspace of 𝐹 if
(i) for each 𝜆 ∈ 𝐴, 𝐺(𝜆) is a vector subspace of 𝑉 over 𝐾 and
(ii) 𝐹(𝜆) ⊇ 𝐺(𝜆), ∀ 𝜆 ∈ 𝐴.
Definition 2.14 (see [7]): Let 𝐺 be a soft vector space of 𝑉 over 𝐾. Then a soft element of 𝐺
is said to be a soft vector of 𝐺. In a similar manner a soft element of the soft set (𝐾, 𝐴) is said
to be a soft scalar, 𝐾 being the scalar field.

Definition 2.15 (see [8]): Let ℂ be the set of complex numbers and 𝜌(ℂ) be the collection of
all non-empty bounded subsets of the set of complex numbers. A be a set of parameters. Then
a mapping 𝐹: 𝐴 → 𝜌(ℂ) is called a soft complex set. It is denoted by (𝐹, 𝐴).

If in particular (𝐹, 𝐴) is a singleton soft set; then identifying (𝐹, 𝐴) with the corresponding
soft element, it will be called a soft complex number. The set of all soft complex numbers is
denoted by ℂ(𝐴).

Definition 2.16 (see [8]): Let (𝐹, 𝐴) be a soft complex set. Then the complex conjugate of
(𝐹, 𝐴) is denoted by (𝐹̅ , 𝐴) and is defined by 𝐹̅ (𝜆) = { 𝑧̅ ∶ 𝑧 ∈ 𝐹(𝜆)} , ∀ 𝜆 ∈ 𝐴 , where 𝑧̅
is complex conjugate of the ordinary complex number 𝑧 . The complex conjugate of a soft
complex number (𝐹, 𝐴) is 𝐹̅ (𝜆) = 𝑧̅ , ∈ 𝐹(𝜆) , ∀ 𝜆 ∈ 𝐴.

Definition 2.17 (see [8]): Let (𝐹, 𝐴), (𝐺, 𝐴) ∈ ℂ(𝐴) . Then the sum, difference, product and
division are defined by
(𝐹 + 𝐺)(𝜆) = 𝑧 + 𝑤 ; 𝑧 ∈ 𝐹(𝜆) , 𝑤 ∈ 𝐺(𝜆) , ∀ 𝜆 ∈ 𝐴;

(𝐹 − 𝐺)(𝜆) = 𝑧 − 𝑤 ; 𝑧 ∈ 𝐹(𝜆) , 𝑤 ∈ 𝐺(𝜆) , ∀ 𝜆 ∈ 𝐴;

(𝐹𝐺)(𝜆) = 𝑧𝑤 ; 𝑧 ∈ 𝐹(𝜆) , 𝑤 ∈ 𝐺(𝜆) , ∀ 𝜆 ∈ 𝐴;

𝐹
(𝐺) (𝜆) = 𝑧/𝑤 ; 𝑧 ∈ 𝐹(𝜆) , 𝑤 ∈ 𝐺(𝜆) , ∀ 𝜆 ∈ 𝐴; provided 𝐺(𝜆) ≠ 𝜙 , ∀ 𝜆 ∈ 𝐴.

Definition 2.18 (see [8]): Let (𝐹, 𝐴) be a soft complex number. Then the modulus of (𝐹, 𝐴)
is denoted by (|𝐹|, 𝐴) and is defined by |𝐹|(𝜆) = |𝑧|; 𝑧 ∈ 𝐹(𝜆), ∀ 𝜆 ∈ 𝐴 , where 𝑧 is an
ordinary complex number.

Since the modulus of each ordinary complex number is a non-negative real number and by
definition of soft real numbers it follows that (|𝐹|, 𝐴) is a non-negative soft real number for
every soft complex number (𝐹, 𝐴).

Definition 2.19 (see [8]): Let 𝑋̌ be the absolute soft vector space i.e., 𝑋̌(𝜆) = 𝑋 , ∀ 𝜆 ∈ 𝐴.
The a mapping < . > ∶ 𝑆𝐸(𝑋̌) × 𝑆𝐸(𝑋̌) → ℂ(𝐴) is said to be a soft inner product on the soft
vector space 𝑋̌ if < . > satisfies the following conditions:

(I1). < 𝑥̃ , 𝑥̃ > ≥ 0̅ , for all 𝑥̃ ∈

̃ 𝑋̌ and < 𝑥̃ , 𝑥̃ > = 0̅ if and only if 𝑥̃ = Φ ;

(I2). < 𝑥̃, 𝑦̃ > = ̅̅̅̅̅̅̅̅̅̅̅̅

< 𝑦̃, 𝑥̃ > , where bar denote the complex conjugate of soft complex
numbers;

̃ 𝑋̌ and for every soft scalar 𝛼̃;

(I3) < 𝛼̃. 𝑥̃, 𝑦̃ > = 𝛼̃ < 𝑥̃, 𝑦̃ > , for all 𝑥̃ , 𝑦̃ ∈

̃ 𝑋̌, < 𝑥̃ + 𝑦̃ , 𝑧̃ > = < 𝑥̃ , 𝑧̃ > + < 𝑦̃ , 𝑧̃ >.

(I4) For all 𝑥̃ , 𝑦̃ ∈

The soft vector space 𝑋̌ with a soft inner product < . > on 𝑋̌ is said to be a soft inner product
space and is denoted by ( 𝑋̌, < . >, 𝐴) or (𝑋̌ , < . >). (I1), (I2) , (I3) and (I4) are said to be soft
inner product axioms.

Proposition 2.20 (see [8]): Let { < . >𝜆 ∶ 𝜆 ∈ 𝐴 } be a family of crisp inner products on a crisp
vector space 𝑋. Then the mapping < . >∶ 𝑆𝐸(𝑋̌) × 𝑆𝐸( 𝑋̌) → ℂ(𝐴) by < 𝑥̃, 𝑦̃ > (𝜆) = <
̃ 𝑋̃ is a soft inner product on the soft linear space 𝑋̌.
𝑥̃(𝜆), 𝑦̃(𝜆) >𝜆 , ∀ 𝜆 ∈ 𝐴, 𝑥̃ , 𝑦̃ ∈

Corollary 2.21 (see [8]): Every crisp inner product space < . >𝑋 on a crisp vector space 𝑋
can be extended to a soft inner product on the soft vector space 𝑋̌.
Theorem 2.22 (see [8]): If a soft inner product < . > satisfies the condition (I5) , and if for
each 𝜆 ∈ 𝐴, < . >𝜆 ∶ 𝑋 × 𝑋 → ℂ be a mapping such that for all (𝜉, 𝜂) ∈ 𝑋 × 𝑋 , < 𝜉, 𝜂 >𝜆 = <
̃ 𝑋̌ such that 𝑥̃(𝜆) = 𝜉 , 𝑦̃(𝜆) = 𝜂. Then for each 𝜆 ∈ 𝐴 , <. >𝜆 is
𝑥̃ , 𝑦̃ > (𝜆) , where 𝑥̃ , 𝑦̃ ∈
an inner product on 𝑋.

̃ 𝑋̌ such that 𝑥̃(𝜆) = 𝜉 ,

(I5) For each (𝜉, 𝜂) ∈ 𝑋 × 𝑋 and ∈ 𝐴 , { < 𝑥̃ , 𝑦̃ > (𝜆) ∶ 𝑥̃ , 𝑦̃ ∈
𝑦̃(𝜆) = 𝜂 } is a singleton set.

Definition 2.23 (see [8]): A soft inner product space which satisfies (I5), is said to be
complete if it is complete with respect to the soft metric defined by soft inner product. A
complete soft inner product space is said to be a soft Hilbert space.

Definition 2.24 (see [8]): Let 𝐿̃ be a non-null soft subset of the soft Hilbert space 𝐻
̃ such that

𝐿̃(𝜆) ≠ 𝜙 , ∀ 𝜆 ∈ 𝐴. Two soft vectors 𝑥̃ , 𝑦̃ of 𝐻

̃ are said to be orthogonal if < 𝑥̃ , 𝑦̃ > = 0̅. In
symbol, we write 𝑥̃ ⊥ 𝑦̃ . If 𝑥̃ is orthogonal to every soft vectors of 𝐿̃ then we say that 𝑥̃ is
orthogonal to 𝐿̃ and we write 𝑥̃ ⊥ 𝐿̃.

̃ be a Hilbert space. Then a collection ℬ of soft vectors of 𝐻

Definition 2.25 (see [8]): Let 𝐻 ̃
is said to be orthonormal if for all 𝑥̃ , 𝑦̃ ∈ ℬ

0̅ ; 𝑖𝑓 𝑥̃ ≠ 𝑦̃
< 𝑥̃ , 𝑦̃ > = {
1̅ ; 𝑖𝑓 𝑥̃ = 𝑦̃

If the soft set ℬ contains only a countable number of soft vectors then we can arrange it in a
sequence of soft vectors and call it an orthonormal sequence.

Definition 2.26 (see [7]): Let G be a soft vector space of V over K. Let 𝛼
̃1 , 𝛼
̃2 , … , 𝛼
̃𝑛 ∈ 𝐺.
A soft vector 𝛽̃ is said to be a linear combination of the soft vectors 𝛼
̃1 , 𝛼 ̃𝑛 if 𝛽̃ can be
̃2 , … , 𝛼
expressed as a 𝛽̃ = 𝑐̃1 . 𝛼
̃1 + 𝑐̃2 𝛼 𝑐𝑛 ̃𝑛 , for some soft scalars 𝑐̃1 , 𝑐̃2 , … , ̃
̃2 + ⋯ + ̃𝛼 𝑐𝑛 .

Definition 2.27 (see [7]): A finite set of soft vectors { 𝛼

̃1 , 𝛼 ̃𝑛 } of soft vector space G
̃2 , … , 𝛼
𝑐𝑛 not all 0̅ such
is said to be linearly independent in G if there exists soft scalars 𝑐̃1 , 𝑐̃2 , … , ̃
that 𝑐̃1 . 𝛼
̃1 + 𝑐̃2 𝛼 𝑐𝑛 ̃𝑛 = Φ . An arbitrary set S of soft vectors of G is said to be
̃2 + ⋯ + ̃𝛼
linearly dependent in G if there exists a finite subset of S which is linearly dependent in G.

Theorem 2.28 (see [7]): An orthonormal set of soft vectors is linearly independent.

3. Gram-Schmidt Orthogonalization Process

Let X be a vector space over a field ℂ of complex number, X is our initial universal set and A
be a non-empty set of parameters. Let 𝑋̌ be the absolute soft vector space i.e. 𝑋̌(𝜆) = 𝑋 . for
all 𝜆 ∈ 𝐴 . We use the notation 𝑥̃ , 𝑦̃ , 𝑧̃ to denote soft vectors of a soft vector space and 𝑟̃ , 𝑠̃ , 𝑡̃
to denote soft real numbers whereas 𝑟̅ , 𝑠̅ , 𝑡̅ will denote a particular type of soft real numbers
such that 𝑟̅ (𝜆) = 𝑟 , for all 𝜆 ∈ 𝐴. For example 0̅ is the soft real number where 0̅(𝜆) = 0 , for
all 𝜆 ∈ 𝐴.

Definition 3.1 (Basis of soft linear space): Let 𝑋̌ be a soft linear space. A set 𝑆 of soft vectors
in 𝑋̌ is said to be a basis of 𝑋̌ if 𝑆 is linearly independent and 𝑆 generates 𝑋̌ i.e. any soft
element of 𝑋̌ can be expressed as a linear combination of those linearly independent soft
vectors.

Definition 3.2 ( Finite dimensional soft linear space): If the basis set 𝑆 of soft vectors is finite
then 𝑋̌ is said to be a finite dimensional soft linear space and the number of soft vectors of the
basis is called the dimension of the soft linear space 𝑋̌.

̃ be a Hilbert space. Then a basis 𝑆 of soft

Definition 3.3 (Soft orthonormal basis): Let 𝐻
̌ is soft orthonormal basis if the set 𝑆 is orthonormal i.e. for all 𝑥̃ , 𝑦̃ ∈ 𝑆 ,
vectors of 𝐻

0̅ ; 𝑖𝑓 𝑥̃ ≠ 𝑦̃
< 𝑥̃ , 𝑦̃ > = {
1̅ ; 𝑖𝑓 𝑥̃ = 𝑦̃

Definition 3.4 (Projection): Let 𝛽̃ be a fixed non-null soft vector in a Hilbert space 𝐻
̌ . Then
̌ there exists a unique soft real number 𝑐̃ such that 𝛼̃ − 𝑐̃ 𝛽̃ is
for a non-null soft vector 𝛼̃ in 𝐻
orthogonal to 𝛽̃. 𝑐̃ is determined by the relation < 𝛼̃ − 𝑐̃ 𝛽̃ , 𝛽̃ > = 0 . Therefore < 𝛼̃ , 𝛽̃ > =
̃>
̃ ,𝛽
<𝛼
𝑐̃ < 𝛽̃ , 𝛽̃ > , giving 𝑐̃ = <𝛽̃,𝛽̃> .

𝑐̃ is said to be the scalar component of 𝛼̃ along 𝛽̃ and 𝑐̃ 𝛽̃ is said to be the projection of 𝛼̃

upon 𝛽̃.

Gram-Schmidt Orthogonalization Process on Euclidean space (Review): Let

{ 𝑥1 , 𝑥2 , … , 𝑥𝑛 } be a basis for n-dimensional inner product space V. An orthonormal basis
will be obtained by the method of Gram-Schmidt process of orthogonalisation.

Let
 𝑥1
𝑢1 =
‖𝑥1 ‖

and

𝑥2 − 𝑐1 𝑢1
𝑢2 =
‖𝑥2 − 𝑐1 𝑢1 ‖

where 𝑐1 𝑢1 is the projection of 𝑥2 upon 𝑢1 i.e.

𝑥2 −< 𝑢1 , 𝑥2 > 𝑢1
𝑢2 =
‖𝑥2 −< 𝑢1 , 𝑥2 > 𝑢1 ‖

Of course, 𝑥2 − < 𝑥2 , 𝑢1 > 𝑢1 ≠ 0 , because {𝑥1 , 𝑥2 } linearly independent . Generally one

can define by induction on 𝑘 = 1 , 2, … , 𝑛,

𝑥𝑘 −< 𝑢1 , 𝑥𝑘 > 𝑢1 −< 𝑢2 , 𝑥𝑘 > 𝑢2 − ⋯ −< 𝑢𝑘−1 , 𝑥𝑘 > 𝑢𝑘−1

𝑢𝑘 =
‖𝑥𝑘 −< 𝑢1 , 𝑥𝑘 > 𝑢1 −< 𝑢2 , 𝑥𝑘 > 𝑢2 − ⋯ −< 𝑢𝑘−1 , 𝑥𝑘 > 𝑢𝑘−1 ‖

Then it is elementary to check that the vectors 𝑢1 , 𝑢2 , … , 𝑢𝑛 are orthonormal in the n-

dimensional vector space V. Since every orthonormal set is linearly independent , it is an
orthonormal basis for V.

Gram-Schmidt Orthogonalization Process on soft inner product space : Let

{x̃1 , x̃,
2 …,x
̃ ̌
n } be a basis of the finite dimensional soft Hilbert space 𝐻 . An orthogonal basis

̌ will be obtained by the Gram-Schmidt process of orthogonalisation. Since the basis

of 𝐻
vectors are non-null, we select one of them , say x̃1 , and consider as the first member of the
new basis. For convenience we rename it as ỹ1 i.e. ỹ1 = x̃1 .

Now let ỹ2 = x̃2 − 𝑐̃1 ỹ1 , where 𝑐̃1 ỹ1 is the projection of ̃
x2 upon ỹ1 .

Then < ỹ2 , ỹ1 > = < x̃2 − 𝑐̃1 ỹ1 , ỹ1 > = < x̃2 , ỹ1 > −c̃1 < ỹ1 , ỹ1 >

<x̃,ỹ>
2 1
= < x̃2 , ỹ1 > − <ỹ,y
̃>
< ỹ1 , ỹ1 > = 0̅.
1 1

This gives ̃
y2 is orthogonal to x̃1 and {ỹ1 , ̃
y2 } = 𝐿 {ỹ1 , ̃ x2 , where 𝐿{ỹ1 , ỹ2 }
x2 } = 𝐿{x̃1 , ̃}
denote the linear span of the soft vectors ỹ1 and ỹ2 .

< ̃,
x2 ỹ1 >
y2 = x̃2 −
̃ ỹ
< ỹ1 , ỹ1 > 1
Now clearly x̃3 ∉ 𝐿{ỹ1 , ̃ ̃1 ỹ1 − d
y2 } . Let ỹ3 = x̃3 − d ̃2 ỹ2 , where d
̃1 ỹ1 , d
̃2 ỹ2 are the
projections of ̃
x3 upon ỹ1 , ̃
y2 respectively.

Then, < ỹ3 , ̃

y2 > = < ̃ ̃1 ỹ1 − d
x3 − d ̃2 ̃
y2 , ỹ2 >

̃1 < ỹ1 , ỹ2 > −d

= < x̃3 , ỹ2 > −d ̃2 < ̃,
y2 ̃
y2 >

̃2 < ̃,
=< x̃3 , ỹ2 > −d y2 > = 0̅
y2 ỹ2 > , since < ỹ1 , ̃

<x̃,ỹ>
3 2
= < x̃3 , ỹ2 > − <ỹ,y
̃>
< ̃,
y2 ỹ2 >
2 2

=< x̃3 , ỹ2 > −< x̃3 , ỹ2 > = 0̅

In a similar way one can check that < ỹ3 , ỹ1 > = 0̅ .

This gives ̃
y3 is orthogonal to ỹ, y1 and 𝐿{ỹ1 , ̃
2 ̃ y3 = 𝐿 {ỹ1 , ̃
y2 , ̃} y2 , x̃3 } = 𝐿{x̃1 , ̃,
x2 ̃}.
x3

< x̃,
3 ỹ1 > < ̃,
x3 ỹ2 >
ỹ3 = ̃
x3 − ỹ1 − ỹ
< ỹ1 , ỹ1 > y2 ỹ2 > 2
< ̃,

Now x̃4 ∉ 𝐿{ỹ1 , ̃ y3 } . Let ỹ4 = ̃

y2 , ̃ x4 − r̃1 ỹ1 − r̃2 ̃ y3 , where r̃1 ỹ1 , r̃2 ̃
y2 − r̃3 ̃ y2 , r̃3 ỹ3
are the projections of ̃
x4 upon ỹ1 , ̃
y2 , ỹ3 respectively.

As proceeding above we can show that ̃

y4 is orthogonal to ỹ1 , ỹ2 , ỹ3 and 𝐿{ỹ1 , ỹ2 , ỹ,
3 ̃}
y4 =
𝐿 {ỹ1 , ỹ, ̃3 , x̃4 } = 𝐿{x̃1 , x̃,
2 𝑦 2 ̃,
x3 x̃}.
4

< ̃,
x4 ỹ1 > < x̃,
4 ̃
y2 > < x̃,
4 ỹ1 >
y4 = ̃
̃ x4 − ỹ1 − y2 −
̃ ỹ
< ỹ1 , ỹ1 > < ỹ,
2 ̃
y2 > < ỹ1 , ỹ1 > 3

This process concludes after a finite number of steps because, at each step, a vector from the
original basis is replaced with a vector from the targeted orthogonal basis. Thus we have,

< x̃,
n ỹ1 > < x̃,
n ỹ2 > < x̃,
n y
̃n−1 >
ỹn = x̃n − ỹ1 − ỹ2 − ⋯ − ỹ
n−1
< ỹ1 , ỹ1 > < ̃,
y2 ỹ2 > < ỹ,
n−1 ỹ
n−1 >

and {ỹ1 , ̃, ̌
n } is an orthogonal basis of the soft Hilbert space 𝐻 .
y2 … , ỹ

Finally from this orthogonal basis one can easily get the orthonormal basis for the soft Hilbert
̌ . The orthonormalization process discussed above by which we can get a orthonormal
space 𝐻
basis from a given basis of soft Hilbert space is known as Gram-Schmidt Orthogonalization
Process on soft inner product space.
4. Conclusion:
In this paper we have studied the Gram-Schmidt Orthogonalization Process, which allow us
to transform a set of linearly independent soft vectors into a set of orthonormal soft vectors,
by analysing the orthogonality of soft vectors in a soft Hilbert space. Furthermore by Gram-
Schmidt orthogonalization process one can easily check that the soft Hilbert space always
possess an orthonormal basis. There is a plenty scope for further research on applications of
this orthogonalization process on soft Hilbert space.

References:

[1] H. Aktas and N. Cagman, Soft sets and soft groups, Inform. Sci. 177 (2007) 2226-
2735.
[2] M. I. Ali, F. Feng, X. Liu, W. K. Min and M. Shabir, On some new operations in soft
set theory, Comput. Math. Appl. 57 (2009) 1547-1553.
[3] D. Chen, The parametrization reduction of soft sets and its applications, Comput.
Math. Appl. 49 (2005) 757-763.
[4] S. Das and S. K. Samanta, Soft real sets, soft real numbers and their properties, J.
Fuzzy Math. 20(3) (2012) 551-576.
[5] S. Das and S. K. Samanta, On soft complex sets and soft complex numbers, J. Fuzzy
Math. 21(1) (2013) 195-216.
[6] S. Das and S. K. Samanta, On soft metric spaces, J. Fuzzy Math. Accepted.
[7] S. Das, Pinaki Majumdar and S. K. Samanta, On soft linear spaces and soft Normed
linear spaces, Communicated to Ukrainian Math. J.
[8] S. Das and S. K. Samanta , On soft inner product spaces , Ann. Fuzzy Math. Inform. 6
(2013), No. 1, 151-170.
[9] F. Feng, Y. B. Jun and X. Zhao, Soft semirings, Comput. Math. Appl. 56 (2008)
2621-2628.
[10] F. Feng, C. X. Li, B. Davvaz and M. I. Ali, Soft sets combined with fuzzy sets
and rough sets: a tentative approach, Soft Computing 14 (2010) 8999-9911.
[11] Y. B. Jun, Soft BCK/BCI-algebras, Comput. Math. Appl. 56 (2008) 1408-
1413.
[12] Z. Kong, L. Gao, L. Wong and S. Li, The normal parameter reduction of soft
sets and its algorithm, Comput. Math. Appl. 56 (2008) 3029-3037.
[13] P. K. Maji, R. Biswas and A. R. Roy, Soft set theory, Comput. Math. Appl.
45(2003) 555 − 562.
[14] P. K. Maji, R. Biswas and A. R. Roy, An application of soft sets in a decision
making problem, Comput. Math. Appl. 44(2002) 1077-1083.
[15] P. Majumdar and S. K. Samanta, On soft mappings, Comput. Math. Appl. 60
(2010) 2666-2672.
[16] D. Molodtsov, Soft set theory first results, Comput. Math. Appl. 37(1999)
19 − 31.
[17] D. Pie and D. Miao, From soft sets to information systems, Granular
Computing, 2005, IEEE International Conference, Volume 2, 617-622.
[18] M. Shabir and M. Naz, On soft topological spaces, Comput. Math. Appl. 61
(2011) 1786-1799.
[19] Y. Zou and Z. Xiao, Data analysis approaches of soft sets under incomplete
information, Knowledge-Based Systems 21 (2008) 941-945.