Paper Implementation - Mendel, Jerry M., and RI Bob John. 'Type-2 fuzzy sets made simple.'

Wed October 27, 2021
type2-fuzzy paper-workout type2-fuzzy-library fuzzy python

Introduction

“Type-2 Fuzzy Sets made simple” is possibly the best paper to learn about Type-2 fuzzy sets and logic. It outlines all the definitions and concepts that are necessary to work with type-2 fuzzy sets in a clear and concise manner. This paper illustrates the implementation of all the examples prepared by Mendel and John using the type2fuzzy library.

This post is the first in a series aimed to illustrate the capabilities of the Type2FuzzyLibrary (https://pypi.org/project/type2fuzzy/) This is achieved by working up the numerical examples in selected papers using the library and comparing the results with those obtained by the original authors. The papers will list the code used to carry out the examples, and the results obtained. All code is written using the Python language.

Type-2 fuzzy set definition

The paper illustrates several type-2 fuzzy sets concepts with a simple general type-2 fuzzy set,

(0.9/0 + 0.8/0.2+ 0.7/0.4 + 0.6/0.6 + 0.5/0.8)/1
(0.5/0 + 0.35/0.2 + 0.35/0.4 + 0.2/0.6 + 0.5/0.8)/2
(0.35/0.6 + 0.35/0.8)/3
(0.1/0 + 0.35/0.2 + 0.5/0.4 + 0.1/0.6 + 0.35/0.8)/4
(0.35/0 + 0.5/0.2 + 0.1/0.4 + 0.2/0.6 + 0.2/0.8)/5

This set will be used in this exercise as in the paper.

$$\tilde{A}=\int_{x\in X}\int_{u\in J_{x}} \mu_{\tilde{A}}(x,u) / (x,u)$$

where $J_{x}\subseteq[0,1]$

The following code snippet illustrates how a general type-2 fuzzy set is defined and used, as explained in Example 1 of the original paper.

Example 1 : definition of the general type-2 fuzzy set

gt2fs_rep =   ''' (0.9/0 + 0.8/0.2+ 0.7/0.4 + 0.6/0.6 + 0.5/0.8)/1
+(0.5/0 + 0.35/0.2 + 0.35/0.4 + 0.2/0.6 + 0.5/0.8)/2
+(0.35/0.6 + 0.35/0.8)/3
+(0.1/0 + 0.35/0.2 + 0.5/0.4 + 0.1/0.6 + 0.35/0.8)/4
+(0.35/0 + 0.5/0.2 + 0.1/0.4 + 0.2/0.6 + 0.2/0.8)/5'''

# create set
gt2fs = GeneralType2FuzzySet.from_representation(gt2fs_rep)

print(f'\nSet representation: {gt2fs}')
Set representation:
(0.9000 / 0.0000 + 0.8000 / 0.2000 + 0.7000 / 0.4000 + 0.6000 / 0.6000 + 0.5000 / 0.8000) / 1.000 +
(0.5000 / 0.0000 + 0.3500 / 0.2000 + 0.3500 / 0.4000 + 0.2000 / 0.6000 + 0.5000 / 0.8000) / 2.0000 +
(0.3500 / 0.6000 + 0.3500 / 0.8000) / 3.0000 + 
(0.1000 / 0.000 + 0.3500 / 0.2000 + 0.5000 / 0.4000 + 0.1000 / 0.6000 + 0.3500 / 0.8000) / 4.0000 +
(0.3500 / 0.0000 + 0.5000 / 0.2000 + 0.1000 / 0.4000 + 0.2000 / 0.6000 + 0.2000 / 0.8000) / 5.000

Vertical Slice

A vertical slice is Type-1 fuzzy set $\mu_{\tilde{A}}(x=x',u)$ for $x\in X$ and $\forall u \in J_{x'}\subseteq[0,1]$, that is:

$$\mu_{\tilde{A}}(x=x',u)=\int_{u\in J_{x'}}f_{x'}(u) / u$$

where $0\leq f_{x'}(u)\leq 1$

The following code snippet illustrates two methods by which a vertical slice can be obtained to replicate the second part of Example 1.

# different ways to get vertical slice
print('mu_a_tilde(',1,')= ', gt2fs.vertical_slice(1))
print('mu_a_tilde(',2,')= ', gt2fs[2])
print('mu_a_tilde(',3,')= ', gt2fs.vertical_slice(3))
print('mu_a_tilde(',4,')= ', gt2fs[4])
mu( 1 )= 0.900/0.000 + 0.800/0.200 + 0.700/0.400 + 0.600/0.600 + 0.500/0.800}
mu( 2 )=0.500/0.000 + 0.350/0.200 + 0.350/0.400 + 0.200/0.600 + 0.500/0.800}
mu( 3 )=0.350/0.600 + 0.350/0.800}
mu( 4 )=0.100/0.000 + 0.350/0.200 + 0.500/0.400 + 0.100/0.600 + 0.350/0.800}

Primary Membership

The domain of a secondary membership function is called the primary membership of $x$. Hence in

$$\tilde{A}=\int_{x \in X} \mu_{ \tilde{A} }(x) /x = \int_{x \in X} \left[ \int_{u\in J_{x}} f(u) / u \right] /x $$

$J_{x}$ is the primary membership function, where $J_{x} \subseteq [0,1]$ for $\forall x \in X$

The code below illustrates the final part of Example 1 where the primary memberships of the general type-2 fuzzy set are listed:


# get the primary memberships of the set
# example 1 (continued)
print('\nPrimary Membership:')
for x_k in gt2fs.primary_domain():
    print('J_',x_k, ' : ',  gt2fs.primary_membership(x_k)) 
Primary Membership:
J1.0  :  [0.0, 0.2, 0.4, 0.6, 0.8]
J2.0  :  [0.0, 0.2, 0.4, 0.6, 0.8]
J3.0  :  [0.6, 0.8]
J4.0  :  [0.0, 0.2, 0.4, 0.6, 0.8]
J5.0  :  [0.0, 0.2, 0.4, 0.6, 0.8]

Secondary Grade

The amplitude of a secondary membership function is the secondary grade. Hence in

$$\tilde{A}=\int_{x \in X} \mu_{ \tilde{A} }(x) /x = \int_{x \in X} \left[ \int_{u\in J_{x}} f(u) / u \right] /x ]$$

where $J_{x} \subseteq [0,1]$, $f(u)$ is the secondary grade.

The following code illustrates the retrieval of selected secondary grade values from the general type-2 fuzzy set

# get the secondary grade of some values
# example 1 (continued)
print('\nSecondary grade of some points:')
print('mu(1,0.2)=',  gt2fs.secondary_grade(1, 0.2), '-- should be 0.8') 
print('mu(2,0)=',  gt2fs.secondary_grade(2, 0), '-- should be 0.5') 
print('mu(3,0.8)=',  gt2fs.secondary_grade(3, 0.8), '-- should be 0.35') 
print('mu(4,0.4)=',  gt2fs.secondary_grade(4, 0.4), '-- should be 0.5') 
Secondary grade of some points:
mu(1,0.2)= 0.8 -- should be 0.8
mu(2,0)= 0.5 -- should be 0.5
mu(3,0.8)= 0.35 -- should be 0.35
mu(4,0.4)= 0.5 -- should be 0.5

Footprint of Uncertainty

The 2D support of $\mu$ is called the footprint of uncertainty (FOU).

$$ FOU(\tilde{A})= \left\lbrace (x,u) \in X \times [0,1] | \mu_{\tilde{A}}(x,u) > 0 \right\rbrace $$

FOU represents the uncertainty in the primary memberships of $\tilde{A}$. It is the union of all primary memberships

$$FOU(\tilde{A}) = \bigcup\limits_{x\in X} J_{x}$$

The FOU can be retrieved using a single line of type2fuzzy library code;

# get the footprint of uncertainty for the set

footprint = gt2fs.footprint_of_uncertainty()
print('\nFootprint of uncertainty: ', footprint)
Footprint of uncertainty:
1.0: CrispSet([0.00000, 0.80000]),
2.0: CrispSet([0.00000, 0.80000]),
3.0: CrispSet([0.60000, 0.80000]),
4.0: CrispSet([0.00000, 0.80000]),
5.0: CrispSet([0.00000, 0.80000])

Embedded Type-2 Fuzzy Sets

For discrete universes of discourse $X$ and $U$, an embedded type-2 set $\tilde{A_e}$ has $N$ elements, where $\tilde{A_e}$ has exactly one element from $J_{x_{1}}, J_{x_{2}}, \dots , J_{x_{N}}$; namely $u_{1}, u_{2}, \dots , u_{N}$ each with associated grade namely $f_{x_{1}}(u_1), f_{x_{2}}(u_2), \dots , f_{x_{N}}(u_N)$, such that:

$$ \tilde{A_e} = \displaystyle \sum_{i=1}^{N} \left[ f_{x_{i}} (u_{i}) \right] / x_{i}$$

where $u_{i} \in J_{x_{i}} \subseteq [0,1]$

Set $\tilde{A_e}$ is embedded in $\tilde{A}$ and there are a total of:

$$Num(\tilde{A_e}) = \displaystyle \prod_{i=1}^{N} M_i$$

In Example 2, the authors depict one of the possible 1250 embedded type-2 fuzzy sets that are possible from the general type-2 fuzzy set. Two examples are presented here:

# number of embedded sets

print('\nNumber of embedded type-2 sets: ',
gt2fs.embedded_type2_sets_count())
\small Number of embedded type-2 sets:  1250
# Example 2

# list all embedded sets

count = 0
print('\nShowing first 10 embedded sets:')
for embedded_set in gt2fs.embedded_type2_sets():
    print(embedded_set)
    count = count+1
    if count > 10:
        break
Showing first 10 embedded sets:
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.1, 0.0, 4.0), (0.35, 0.0, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.1, 0.0, 4.0), (0.5, 0.2, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.1, 0.0, 4.0), (0.1, 0.4, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.1, 0.0, 4.0), (0.2, 0.6, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.1, 0.0, 4.0), (0.2, 0.8, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.35, 0.2, 4.0), (0.35, 0.0, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.35, 0.2, 4.0), (0.5, 0.2, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.35, 0.2, 4.0), (0.1, 0.4, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.35, 0.2, 4.0), (0.2, 0.6, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.35, 0.2, 4.0), (0.2, 0.8, 5.0)]
[(0.9, 0.0, 1.0), (0.5, 0.0, 2.0), (0.35, 0.6, 3.0), (0.5, 0.4, 4.0), (0.35, 0.0, 5.0)]

In the original paper, Example 3 considers the following general type-2 fuzzy set:

(0.5/0.9)/x_1 + (0.2/0.7)/x_1 + (0.9/0.2)/x_1 + (0.6/0.6)/x_2 + (0.1/0.4)/x_2

For the sake of this exercise, we assign the values of $x_1 = 1$ and $x_2=2$

thus obtaining the following set;

(0.5/0.9)/1 + (0.2/0.7)/1 + (0.9/0.2)/1 + (0.6/0.6)/2 + (0.1/0.4)/2
(0.5/0.9)/1 + (0.2/0.7)/1 + (0.9/0.2)/1 + (0.6/0.6)/2 + (0.1/0.4)/2

The embedded type-2 fuzzy sets are listed using the code below;

# Example 3

print('\nEmbedded set listing for general type-2 fuzzy set')
print(str(gt2fs_2))
for embedded_set in gt2fs_2.embedded_type2_sets():
    print(embedded_set)
Embedded set listing for general type-2 fuzzy set
(0.5000 / 0.9000 + 0.2000 / 0.7000 + 0.9000 / 0.2000) / 1.0000 
+ (0.6000 / 0.6000 + 0.1000 / 0.4000) / 2.0000
[(0.9, 0.2, 1.0), (0.1, 0.4, 2.0)]
[(0.9, 0.2, 1.0), (0.6, 0.6, 2.0)]
[(0.2, 0.7, 1.0), (0.1, 0.4, 2.0)]
[(0.2, 0.7, 1.0), (0.6, 0.6, 2.0)]
[(0.5, 0.9, 1.0), (0.1, 0.4, 2.0)]
[(0.5, 0.9, 1.0), (0.6, 0.6, 2.0)]



Paper Implementation - Uncertain rule-based fuzzy logic systems Introduction and new directions-Jerry M. Mendel; Prentice-Hall, PTR, Upper Saddle River, NJ, 2001,    555pp., ISBN 0-13-040969-3. Example 9-4, page 261

October 8, 2022
type2-fuzzy type2-fuzzy-library fuzzy python IT2FS paper-workout

Type Reduction of Interval Type-2 Fuzzy Sets

April 6, 2022
type2-fuzzy type2-fuzzy-library fuzzy python IT2FS

Paper Implementation - C. Wagner and H. Hagras. 'Toward general type-2 fuzzy logic systems based on zSlices.'

A look at C. Wagner and H. Hagras. 'Toward general type-2 fuzzy logic systems based on zSlices.', working of paper examples using T2Fuzz Library
type2-fuzzy paper-workout type2-fuzzy-library fuzzy python
comments powered by Disqus


machine-learning 27 python 21 fuzzy 14 azure-ml 11 hugo_cms 11 linear-regression 10 gradient-descent 9 type2-fuzzy 8 type2-fuzzy-library 8 type1-fuzzy 5 cnc 4 dataset 4 datastore 4 it2fs 4 excel 3 paper-workout 3 r 3 c 2 c-sharp 2 experiment 2 hyperparameter-tuning 2 iot 2 model-optimization 2 programming 2 robotics 2 weiszfeld_algorithm 2 arduino 1 automl 1 classifier 1 computation 1 cost-functions 1 development 1 embedded 1 fuzzy-logic 1 game 1 javascript 1 learning 1 mathjax 1 maths 1 mxchip 1 pandas 1 pipeline 1 random_walk 1 roc 1 tools 1 vscode 1 wsl 1