International Journal of Science, Engineering and Technology Research (IJSETR)
Volume x, Issue x, July 2012
DEVELOPMENT OF A PREDICTIVE FUZZY EXPERT SYSTEM
FOR DIAGNOSING CASSAVA PLANT DISEASES.
Adebisi, R. O, Awoyelu, I. O and Adegoke, B. O.

Abstract
This paper presents a fuzzy expert system for predicting
outbreak of some cassava plant diseases. Cassava being a
very important tropical root crop that is capable of providing
starch for use in drug industries and, a stable source of dietary
energy for more than 500 million (FAO). The fuzzy expert
system predicts accurately once the favorable conditions are
measured and quantized. The system was developed with the
help of fuzzy tool in MATLAB vs. 9. It employed 243 rules
for its classification and prediction. The result from the
RSME value showed that the system was able to effectively
predict and classify cassava diseases considered.
Index Terms — Cassava Mosaic disease (CMD), Artificial
Neural Network (ANN),Radial Basis Neural Network
(RBNN), mean square error (MSE).
I.
INTRODUCTION
Cassava is the third largest source of carbohydrates for
human consumption worldwide, providing more food
calories per cultivated acre than any other staple crop. It is an
extremely robust plant which tolerates drought and low
quality soil. The foremost cause of yield loss for this crop is
viral disease (Otim-Nape, Alicai, and Thresh 2005). A major
factor keeping East African farmers trapped in poverty (The
Economist, 2011). Cassava is the West Indian name and is
used in the United States; manioc, or mandioc, is the
Brazilian name; and juca, or yucca, is used in other parts of
South America. The plant grows in a bushy form, up to 2.4 m
(up to 8 ft) high, with greenish-yellow flowers. The roots are
up to 8 cm (up to 3 in) thick and 91 cm (36 in) long.
Two varieties of the cassava are of economic value: the
bitter, or poisonous; and the sweet, or non-poisonous,
because the volatile poison can be destroyed by heat in the
process of preparation, both varieties yield a wholesome
food. Cassava is the chief source of tapioca, and in South
America a sauce and an intoxicating beverage are prepared
from the juice. The root in powder form is used to prepare
farinha, a meal used to make thin cakes sometimes called
cassava bread. The starch of cassava yields a product called
.
First Author name, His Department Name, University/ College/
Organization Name, ., (e-mail: fisrtauthor@gamil.com). City Name,
Country Name, Phone/ Mobile No
Second Author name, His Department Name, University/ College/
Organization Name, City Name, Country Name, Phone/ Mobile No., (e-mail:
secondauthor@rediffmail.com).
Third Author name, His Department Name, University/ College/
Organization Name, City Name, Country Name, Phone/ Mobile No., (e-mail:
thirdauthor@hotmail.com).
Brazilian arrowroot. In Florida, where sweet cassava is
grown, the roots are eaten as food, fed to stock, or used in the
manufacture of starch and glucose.
The economies of many developing countries are
dominated by an agricultural sector in which small-scale and
subsistence farmers are responsible for most production,
utilizing relatively low levels of agricultural technology. As a
result, disease among staple crops presents a serious risk,
with the potential for devastating consequences. It is
therefore critical to monitor the spread of crop disease,
allowing targeted interventions and foreknowledge of famine
risk. Currently, teams of trained agriculturalists are sent to
visit areas of cultivation across the country and make
assessments of crop health. A combination of factors
conspire to make this process expensive, untimely and
inadequate, including the scarcity of suitably trained staff, the
logistical difficulty of transport, and the time required to
coordinate paper reports.(Quinn, et.al., 2010).
An expert system is a system that could keep knowledge in
its knowledge base as the system knowledge resources and
manipulate that knowledge, so it could prepare the high level
decision tool to the user that is called inference engine as the
brain of the system. On the other hand, expert system could
help people in many cases in order to get decision in solving a
problem. This work aimed at developing an expert system
for diagnosing and predicting cassava diseases which could
also be used both by the farmer and the experts to train their
students. The knowledge was elicited from the experts
through interview and literature review. The knowledge was
represented in the system using a rule based approach
(Arowolo, et al; 2012).
II. LITERATURE REVIEW
Different types of learning paradigm exists which
include artificial neural network, fuzzy inference system, the
adaptive neuro fuzzy inference system.
The Fuzzy Logic tool was introduced in 1965, by Lotfi
Zadeh, and is a mathematical tool for dealing with
uncertainty; In other words, it offers to a soft computing
partnership which is the important concept of computing with
words’. It provides a technique to deal with imprecision and
information granularity. The fuzzy theory provides a
mechanism for representing linguistic constructs such as
“many,” “low,” “medium,” “often,” “few.” In general, the
fuzzy logic provides an inference structure that enables
appropriate human reasoning capabilities (Sivanandam,
et.al., 2002).
On the contrary, the traditional binary set theory describes
crisp events, events that either do or do not occur. It uses
probability theory to explain if an event will occur,
measuring the chance with which a given event is expected to
1
All Rights Reserved © 2015 IJSETR
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 1, Issue 1, July 2012
occur. The theory of fuzzy logic is based upon the notion of
relative graded membership and so are the functions of
mentation and cognitive processes. The utility of fuzzy sets
lies in their ability to model uncertain or ambiguous. It is
important to observe that there is an intimate connection
between Fuzziness and Complexity. As the complexity of a
task (problem), or of a system for performing that task,
exceeds a certain threshold, the system must necessarily
become fuzzy in nature. Zadeh, originally an engineer and
systems scientist, was concerned with the rapid decline in
information afforded by traditional mathematical models as
the complexity of the target system increased. Due to the
increase of complexity our ability to make precise and yet
significant statements about its behaviour diminishes. Real
world problems (situations) are too complex, and the
complexity involves the degree of uncertainty – as
uncertainty increases, so does the complexity of the problem.
Traditional system modelling and analysis techniques are too
precise for such problems (systems), and in order to make
complexity less daunting we introduce appropriate
simplifications, assumptions, etc. (i.e., degree of uncertainty
or Fuzziness) to achieve a satisfactory compromise between
the information we have and the amount of uncertainty we
are willing to accept.
A fuzzy inference system consists of three components.
First, a rule base contains a selection of fuzzy rules;
secondly, a database defines the membership functions used
in the rules and, finally, a reasoning mechanism to carry out
the inference procedure on the rules and given facts. This
combination merges the advantages of fuzzy system and
a neural network. (Sivanandam, et.al; 2002).
Fuzzy set relationship
Fuzzy sets provide means to model the uncertainty
associated with vagueness, imprecision, and lack of
information regarding a problem or a plant, etc.
Consider the meaning of a “short person.” For an
individual X, the short person may be one whose height is
below 4’’25
For other individual Y, the short person may be one
whose height is below or equal to 3’’90.
This “short” is called as a linguistic descriptor. The
term “short” informs the same meaning to the individuals X
and Y, but it is found that they both do not provide a unique
definition. The term “short” would be conveyed effectively,
only when a computer compares the given height value with
the pre-assigned value of “short.” This variable “short” is
called as linguistic variable, which represents the imprecision
existing in the system.
The uncertainty is found to arise from ignorance,
from chance and randomness, due to lack of knowledge, from
vagueness (unclear), like the fuzziness existing in our natural
language. Lotfi Zadeh proposed the set membership idea to
make suitable decisions when uncertainty occurs. Consider
the “short” example discussed previously. If we take “short”
as a height equal to or less than 4 feet, then 3’’90 would easily
become the member of the set “short” and 4’’25will not be a
member of the set “short.” The membership value is “1” if it
belongs to the set or “0” if it is not a member of the set. Thus
membership in a set is found to be binary i.e., the element is a
member of a set or not.
It can be indicated as,
Where, χA(x) is the membership of element x in set
A and A is the entire set on the universe.
Fuzzy Rules and Analysis
Rules form the basis for the fuzzy logic to obtain the fuzzy
output. The rule based system is different from the expert
system in the manner that the rules comprising the rule-based
system originates from sources other than that of human
experts and hence are different from expert systems. The
rule-based form uses linguistic variables as its antecedents
and consequents. The antecedents express an inference or the
inequality, which should be satisfied. The consequents are
those, which we can infer, and is the output if the antecedent
inequality is satisfied. The fuzzy rule-based system uses
IF–THEN rule-based system, given by, IF antecedent, THEN
consequent. The formation of the fuzzy rules is discussed in
this chapter.
Formation of Rules
The formation of rules is in general the canonical rule
formation. For any linguistic variable, there are three general
forms in which the canonical rules can be formed. They are:
(1) Assignment statements
(2) Conditional statements
(3) Unconditional statements
Assignment Statements
These statements are those in which the variable is
assignment with the value. The variable and the value
assigned are combined by the assignment operator “=.” The
assignment statements are necessary in forming fuzzy rules.
The value to be assigned may be a linguistic term.
The examples of this type of statements are:
y =low,
Sky color = blue,
Climate = hot
A =5
p=q+r
Temperature = high
The assignment statement is found to restrict the
value of a variable to a specific equality.
Conditional Statements
In this type of statements, some specific conditions are
mentioned, if the conditions are satisfied then it enters the
following statements, called as restrictions.
If x = y Then both are equal,
If Mark > 50 Then pass,
If Speed > 1, 500 Then stop.
If Speed > 1, 500 Then stop.
These statements can be said as fuzzy conditional statements,
such as
If condition C Then restriction F
Unconditional Statements
There is no specific condition that has to be satisfied in this
form of statements.
Some of the unconditional statements are:
Go to F/o
Push the value
Stop
2
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume x, Issue x, July 2012
The control may be transferred without any
appropriate conditions. The unconditional restrictions in the
fuzzy form can be:
R1 : Output is B1
AND
R2 : Output is B2
AND
..., etc.
where B1 and B2 are Fuzzy consequents.
Both conditional and unconditional statements place
restrictions on the consequent of the rule-based process
because of certain conditions. The fuzzy sets and relations
model the restrictions. The linguistic connections like “and,”
“or,” “else” connects the conditional, unconditional, and
restriction statements. The consequent of rules or output is
denoted by the restrictions R1, R2,...,Rn. The rule-based
system with a set of conditional rules (canonical form of
rules) is shown in the table below.
Rule 1:
If condition C1 then Restriction R1
Rule 2:
If condition C2 then Restriction R2
Rule 3:
If condition C3 then Restriction R3
Rule n:
If condition Cn then Restriction Rn
Fuzzy Inference System
Fuzzy inference system consists of a fuzzification
interface, a rule base, a database, a decision-making unit, and
finally a de-fuzzification interface. A fuzzy inference system
with five functional block described in figure 2.2.
The function of each block is as follows:
(i)
A rule base containing a number of fuzzy
IF–THEN rules;
(ii)
A
database
which
defines
the
membership
functions of the fuzzy sets used in the fuzzy rules;
(iii)
A decision-making unit which performs
the
inference operations on the rules;
(iv)
A fuzzification interface which transforms
the
crisp inputs into degrees of match with linguistic
values;
(v)
A
defuzzification
interface
which
transforms the
fuzzy results of the inference into a crisp output.
Fuzzy inference systems (FISs) are also known as fuzzy
rule-based systems, fuzzy model, fuzzy expert system, and
fuzzy associative memory. This is a major unit of a fuzzy
logic system. The decision-making is an important part in the
entire system. The FIS formulates suitable rules and based
upon the rules the decision is made. This is mainly based on
the concepts of the fuzzy set theory, fuzzy IF–THEN rules,
and fuzzy reasoning. FIS uses “IF...THEN...” statements, and
the connectors present in the rule statement are “OR” or
“AND” so as to make necessary decision rules. The basic FIS
can take either fuzzy inputs or crisp inputs, but the outputs it
produces are almost always fuzzy sets. When the FIS is used
as a controller, it is necessary to have a crisp output.
Therefore in this case de fuzzification method is adopted to
best extract a crisp value that best represents a fuzzy set.
(Hossan and Ahmad, 2012).
Fig 2.2 A fuzzy inference system (Sivanadam et.al., 2007)
The working of a FIS is as follows. The crisp input is
converted in to fuzzy by using fuzzification method. After
fuzzification the rule base is formed. The rule base and the
database are jointly referred to as the knowledge base.
De-fuzzification is used to convert fuzzy value to the real
world value which is the output. The steps of fuzzy reasoning
(inference operations upon fuzzy IF–THEN rules) performed
by FISs are:
1.
Compare the input variables with the membership
functions on the antecedent part to obtain the
membership values of each linguistic label (this step
is often called fuzzification.)
2.
Combine (through a specific t-norm operator,
usually multiplication ormin) the membership
values on the premise part to get firing strength
(weight) of each rule.
3.
Generate the qualified consequents (either fuzzy or
crisp) or each rule depending on the firing strength.
4.
Aggregate the qualified consequents to produce a
crisp output. (This step is called de-fuzzification)
The most important two types of fuzzy inference
method are Mamdani’s fuzzy inference method, which is the
most commonly seen inference method. This method was
introduced by Mamdani and Assilian (1975). Another
well-known inference method is the so-called Sugeno or
Takagi–Sugeno–Kang method of fuzzy inference process.
This method was introduced by Sugeno (1985).
This method is also called as TS method. The main
difference between the two methods lies in the consequent of
fuzzy rules. Mamdani fuzzy systems use fuzzy sets as rule
consequent whereas TS fuzzy systems employ linear
functions of input variables as rule consequent. All the
existing results on fuzzy systems as universal approximators
deal with Mamdani fuzzy systems only and no result is
available for TS fuzzy systems with linear rule consequent.
2.3.4 Mamdani Fuzzy Inference
Mamdani’s fuzzy inference method is the most
commonly seen fuzzy methodology. Mamdani’s method was
among the first control systems built using fuzzy set theory. It
was proposed by Mamdani (1975) as an attempt to control a
steam engine and boiler combination by synthesizing a set of
linguistic control rules obtained from experienced human
operators. Mamdani’s effort was based on Zadeh’s (1973)
paper on fuzzy algorithms for complex systems and decision
processes.
2.3.5 Takagi–Sugeno Fuzzy Method (TS Method)
The Sugeno fuzzy model was proposed by Takagi,
Sugeno, and Kang in an effort to formalize a system approach
to generating fuzzy rules from an input–output data set.
Sugeno fuzzy model is also known as Sugeno–Takagi model.
A typical fuzzy rule in a Sugeno fuzzy model has the format.
3
All Rights Reserved © 2015 IJSETR
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 1, Issue 1, July 2012
IF x is A and y is B THEN z = f(x, y),
where AB are fuzzy sets in the antecedent; Z = f(x, y) is a
crisp function in the consequent. Usually f(x, y) is a
polynomial in the input variables x and y, but it can be any
other functions that can appropriately describe the output of
the output of the system within the fuzzy region specified by
the antecedent of the rule. When f(x, y) is a first-order
polynomial, we have the first-order Sugeno fuzzy model.
When f is a constant, we then have the zero-order Sugeno
fuzzy model, which can be viewed either as a special case of
the Mamdani FIS where each rule’s consequent is specified
by a fuzzy singleton, or a special case of Tsukamoto’s fuzzy
model where each rule’s consequent is specified by a
membership function of a step function centered at the
constant. Moreover, a zero-order Sugeno fuzzy model is
functionally equivalent to a radial basis function network
under certain minor constraints.
The first two parts of the fuzzy inference process,
fuzzifying the inputs and applying the fuzzy operator, are
exactly the same. The main difference between Mamdani and
Sugeno is that the Sugeno output membership functions are
either linear or constant. A typical rule in a Sugeno fuzzy
model has the form
IF Input 1 = x AND Input 2 = y, THEN
Output is z = ax + by + c.
Where, AB are fuzzy sets in the antecedent; Z = f(x,
y) is a crisp function in the consequent. Usually f(x, y) is a
polynomial in the input variables x and y,
For a zero-order Sugeno model, the output level zis a constant
(a = b =0).The output level zi of each rule is weighted by the
firing strength wi of the rule. For example, for an AND rule
with Input 1 = x and Input 2 = y,the firing strength is
Wi = AndMethod(F1(x), F2(y))
Where F1,2(.) are membership functions for input 1 and
2.The final output of the system is the weighted average of all
rule outputs, computed as
Fuzzy logic has been used to solve numerous
problems ranging from prediction rate to classification rate.
Some of the research work in which fuzzy logic has been
applied has been summarised below.
In the modelling of crop disease and control, fuzzy
logic was used to model and monitor crop diseases in
developing countries of the world. Computer vision
techniques for using camera-enabled mobile devices to make
disease diagnoses directly, allowing reliance on survey
workers with lower levels of training, and hence reducing
survey costs. Specifically, given expert annotated images of
single cassava leaves, we demonstrate classification based on
color and shape information (Quinn et.al., 2012). Models of
crop disease are used for understanding the spread or severity
of an epidemic, predicting the future spread of infection, and
choosing disease management strategies. Common to all of
these problems is the notion of spatial interpolation.
Observations are made at a few sample sites, and from these
we infer the distribution across the entire spatial field of
interest. Standard approaches to this problem reviewed in
(van Maanen and Xu 2003)) include the use of differential
equations, spatial autocorrelation, or kriging (Gaussian
process regression) (Nelson, Orum, and Jaime-Garcia 1999).
Under these schemes, summary statistics such incidence or
severity are interpolated. Note that the observations contain
more information than these models use; under an incidence
regression, we simplify the raw observation.
Furthermore, Kumar, 2013 was able to use fuzzy
logic to be able to predict and diagnose cassava diseases and
conditions using advanced fuzzy mechanism. In this
procedure crisp values are transferred into fuzzy values
through the fuzzification. Advanced fuzzy resolution
mechanism uses predicted value to diagnosis the cassava
disease with five layers, each layer has its own nodes.
(Kumar, 2013).
Also, ANFIS is used as a system to predict the future
actions of the exchange rate. ANFIS proves very simple to
use and relatively fast, some stability problems do arise that
need to be dealt with before ANFIS can be fully utilized to
predict the exchange rate.
The main conclusion is that more investigations need to
bake place before it is possible to use an ANFIS system to
predict the exchange rate. GARCH systems are rarely used as
the only means of estimating time series, but can give very
useful information. Perhaps in conjunction with an ANFIS
system the GARCH could foresee when the ANFIS would be
likely to be going unstable, this way retraining could take
place and re-estimation of the parameters could also be
conducted.
Moreover, ANFIS has been used for surface
roughness prediction model for ball end milling operation.
For the prediction of surface roughness, a feed forward ANN
was used for face milling of aluminum alloy by
Bernardos et al. (2003) high chromium steel (AISI H11)
by Rai et al. (2010) and AISI 420 B stainless steel by Bruni
et al. (2008). Bruni et al. (2008) proposed the analytical
and artificial neural network models. Yazdi and Khorram
(2010) worked for selection of optimal machining
parameters (that is, spindle speed, depth of cut and feed
rate) for face milling operations in order to minimize the
surface roughness and to maximize the material removal
rate using response surface methodology (RSM) and
perceptron neural network.
Munoz-Escalona
and
Maropoulos
(2009)
proposed the radial basis feed forward neural network
model and generalized regression for surface roughness
prediction for face milling of Al 7075-T735. The Pearson
correlation coefficients were also calculated to analyze the
correlation between the five inputs (cutting speed, feed
per tooth, axial depth of cut, chip's width, and chip's
thickness) with surface roughness. Zhanjie et al. (2007)
used radial basis function network to predict surface
roughness and compared with measured values and the result
from regression analysis. Lu and Costes (2008) considered
three variables, that is, cutting speed, depth of cut and
feed rate to predict the surface profile in turning process
using radial basis function (RBF). Experiments have been
carried out by Brecher et al. (2011) after end milling of
steel C45 in order to obtain the roughness data and model of
ANN for surface roughness predictions.(Hossan and Ahmad,
2012). Aderonmu and his research group developed a
neuro-fuzzy system for Animal Feed Formulation (AFF)
4
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume x, Issue x, July 2012
where percentage of each ingredients in poultry feed was
determined by ANN and their different ingredients were
combined with the help of Neuro-fuzzy and network
(Aderonmu et al; 2013)
III. SYSTEM METHODOLOGY
Data for the cassava diseases were gotten from online
datasets in which was used to train the diseases. Crisp values
are transferred into fuzzy values through the fuzzification.
Advanced fuzzy resolution mechanism uses predicted value
to diagnosis the cassava diseases with five layers, each layer
has its own nodes. The proposed mechanism was tested with
the cassava diseases datasets. Advanced Fuzzy Resolution
Mechanism was
developed
using MATLAB.
Defuzzification converts the fuzzy set into crisp values. The
proposed method with predicted value technique can work
more efficiently for diagnosis of cassava disease and also
compared with earlier method using accuracy as metrics.
Diseases and Conditions
Different type of diseases affects the cassava plant, but in
this project, three of such diseases would be placed into
consideration. This include Cassava Mosaic disease, Cassava
Blight disease, and Cassava Bacteria blight.
The condition for which these cassava disease are;
1) Presence of vector
2) Increase light intensity
3) Temperature of the environment
4) High wind
5) Light rainfall
The value for the fuzzy variables are shown below:
Table 1.: Fuzzy variables and their representation.
Fuzzy variables
Representation
of fuzzy
variable
Presence of vector
X1
Increase light intensity
X2
Temperature of the environment
X3
Light rainfall
X4
High wind
X5
The predicted stock
Y
The input set has a corresponding input of the following
sequence, which implies that the inputs of the fuzzy sets are:
x1, x2, x3, x4, x5.
Output of the fuzzy set is y
The input functions of the variables are represented in
the membership function of the corresponding input and
output, in the system represented the data were not too large,
but prediction was accurate to some extent.
The output function is the result generated from the input
function which was necessarily a detailed representation of
the initial programming parameter.
System Algorithm and Specification
Data were input into the corresponding name of the data
type which was translated into real time analysis and
computing, a variable was used to do the searching algorithm
and exhaustion which aided and help in making comparative
check between trained data and tested data. Step 1 begin.
Step1: Input the crisp values for x1,x2,x3,x4,x5.
Step 2: Set Sugeno fuzzy model, with fuzzy if-then rules.
Step 3: Assign fuzzy numbers for each input variables.
Step 4: Output variable is predicted using the linear equation
𝑦 = 𝑎1 𝑥1 + 𝑎2 𝑥2 + 𝑎3 𝑥3 + 𝑎4 𝑥4 + 𝑎5 𝑥5
Parameter is predicted for linear equation using the formula
{(
𝑎𝑖 =
𝑚𝑒𝑎𝑛(𝑦)
𝑚𝑒𝑎𝑛(𝑦)
⁄max(𝑥 )) + (
⁄𝑚𝑒𝑎𝑛(𝑥 ))}
𝑖
𝑖
𝑚𝑒𝑎𝑛(𝑦) ∗ 𝑛𝑜 𝑜𝑓 𝑓𝑎𝑐𝑡𝑜𝑟𝑠
𝑤ℎ𝑒𝑟𝑒 𝑖 = 1 𝑡𝑜 5
Step 5: Generate the rule as
If input 1 = 𝑥1 or Input 2 = 𝑥2 or Input 3 = 𝑥3 or Input 4 =
𝑥4 or Input 5 = 𝑥5 then Output is
𝑦 = 𝑎1 𝑥1 + 𝑎2 𝑥2 + 𝑎3 𝑥3 + 𝑎4 𝑥4 + 𝑎5 𝑥5
Step 6: Stop
IV. RESULT AND DISCUSSION
The fuzzy system is an in-built system that takes in
multiple input into its membership function, these inputs are
gotten from the data- the data which in turn has numeric
values and assigned numbers and variables are being used to
classify and arrange these data. Sequel to this, some of these
data can be classified as categorical or numerical, in which
there is adequate implementation.
The implementation was carried out with Matlab 7.10a,
implemented on a 2-Ghz memory and in which the final
results were computed.
The system represented above is a function of the
training system that has the testing and checking of data. The
data can either be collected from the file or workspace.
Loading a data from the file means inputting the name of the
saved file where the data has been kept.
Adaptive neuro-fuzzy inference system is a fuzzy
inference system implemented in the framework of an
adaptive neural network. By using a hybrid learning
procedure, ANFIS can construct an input-output mapping
based on both human-knowledge as fuzzy if-thenrules
and approximate membership functions from the
stipulated input-output data pairs for neural network
training. This procedure of developing a FIS using the
framework of adaptive neural networks is called an
adaptive neuro fuzzy inference system (ANFIS). There are
two methods that ANFIS learning employs for updating
membership function parameters:
1) Backpropagation for all parameters (a steepest descent
method), and
2) A hybrid method consisting of backpropagation for the
parameters associated with the input membership and least
squares estimation for the parameters associated with the
5
All Rights Reserved © 2015 IJSETR
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 1, Issue 1, July 2012
output membership functions. As a result, the training error
decreases, at least locally.
rainfall has minimal effect on the diseases due to the low
value of its root mean square error.
Fig 2: ANFIS interface showing loaded data
On clicking the generate fis button, the interface below is
shown
Fig. 5: A Graph of Two Input Variable Compared With
The Root Mean Square Error.
The above graph shows the relationship between the
variables and the root mean square, it was discovered that
light rainfall and temperature of the environment has
minimal effect on the diseases due to the low value of its root
mean square error.
Fig 3: Selection of Membership Function.
The member function gbellmf is selected, this is done in
order to enable the input maximise its usage. This in turn
helps to generate a model input.
First Input Variable
These variables which are tried are tried at the first trial
of the input function each of the variable represented and the
effect shown in the graph below
Fig 6: Graph of three input variable against root mean
square error.
The above graph shows the relationship between the
variables and the root mean square, it is discovered that light
rainfall, temperature of the environment and presence of
vector has minimal effect on the diseases due to the low value
of its root mean square error.
Fig 4: Graph showing the Training Circles and the Root
Mean Square Errors with First Input Variable.
The above graph shows the relationship between the
variables and the root mean square, it is discovered that light
6
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume x, Issue x, July 2012
V. CONCLUSION AND RECOMMENDATION
The adaptive neuro fuzzy inference system has broken
the bounds of conventional programming, which is actually a
function of its ability to adapt, adopt, adjust, evaluate, learn
and recognize the relationship, behaviour and pattern of
series of data set administered to it. It is tailored after the
human reasoning and learning mechanism. This enables the
ANFIS to handle and represent more complex problems than
conventional programming. Adaptive Neuro Fuzzy Inference
System (ANFIS), fuzzy logic and expert system has helped in
diagnosing and predicting diseases.
It can be seen that light rainfall and temperature of the
environment has a larger effect on the diseases of cassava
plants, more to say they show a little significance when
compared to the other factors.
Fig 7: Input-Output Surface for Trained ANFIS
We can see some spurious effects at the near-end corner
of the surface. The elevated corner says that the heavier light
rainfall is, the more the diseases.
Fig 8: ANFIS MODEL OUTPUT showing Rules
In the prediction above, the rules used are 243 rules,
these rules were sufficient for the prediction and the analysis.
Fig.9: The Anfis Model Structure
In the above figure, there are five input into the anfis
model which represents the five variables. Thus the five
variables are light rainfall, temperature, presence of vectors,
high wind, increase light intensity.
V RECOMMENDATIONS
In view of this research work, I strongly recommended that
the following be done. Adequate and enough data should be
gotten from database bank which keeps information on
diseases of plants. The curriculum of the institution should be
reviewed to reflect current technological trends. Future work
can be carried out with respect to more variables such as
humidity, moisture content et cetera.
REFERENCES
Aderonmu G.A, Omidiora E.O., Adegoke B.O., and Taiwo
T.A., (2013). Neuro-fuzzy System for Livestock Feed
Formulation (Africa Poultry). The International Journal
of Engineering and Science (IJES), 2(5), 25 – 32.
Alicai T., C.A. Omongo C.A., (2007). Re-emergence of
Cassava Brown Streak Disease in Uganda. Plant disease.
Vol.91 No. 1
Buttarazzi B., Delfrate F., Solimini C., (2006). User-friendly
automatic tool for image classification based on neural
networks. Springer – Science + Business Media.
Demuth H. and Beale M., (1998). Neural Network toolbox.
The Math Works. Inc., Natick, Massachusetts, 01760.
Fadel M., (2007). Date Fruits Classification Using
Probabilistic Neural Networks. Agricultural Engineering
International: the CIGR E-journal. Vol. IX.
Griner P.F., Mayemski R.J., Mushlin A.I., Greenland P.,
(1981) Selection and interpretation of diagnostic tests
and procedures. Annals of Internal Medicine, 94,
555-600.
Halimat Arowolo, et.al., (2012): “An Expert System for
Management of Poultry Diseases”: International
Conference on Computer Technology and Science
(ICCTS 2012), IPCSIT vol. 47 (2012) IACSIT Press,
Singapore.
Hossain S.J and Ahmad N. (2012): “Adaptive neuro-fuzzy
inference system (ANFIS) based surface roughness
prediction model for ball end milling operation”; Journal
of Mechanical Engineering Research Vol. 4(3), pp.
112-129.
International Institute of Tropical Agriculture, (2003).
Preemptive management of the virulent cassava mosaic
disease through an integrated cassava development
approach for enhanced rural sector economy in the south
–south and south –east zones of Nigeria, pp. 1-79, ISBN
978 131.
7
All Rights Reserved © 2015 IJSETR
International Journal of Science, Engineering and Technology Research (IJSETR)
Volume 1, Issue 1, July 2012
John A. Quinn et.al., (2011): “ Modeling and Monitoring
Crop Disease in Developing Countries”: Association for
the Advancement of Artificial Intelligence.
Kumar M., Bandyopadhyay A., Raghuwanshi N.S., Singh R.,
(2008). Comparative study of conventional and artificial
neural network –based to estimation models. Springer
–Verlag, 126 -128, May 2008.
Ma L., Khorasani K., (2004). New training strategies for
constructive neural networks with application. Neural
Network, no. 17, pp. 589 -609, METZ C.E., (1978) Basic
principles of ROC analysis. Seminars in Nuclear
Medicine, 8, 283-298.
Martin T. Hagan, Howard B. Demuth, Mark Beale, (1996).
Neural Network Design, copyright 1996 by PWS
Publishing Company.
Mucherino A., Petraq Papajorgi, Pardalos P.M, (2009). A
survey of data mining techniques applied to Agriculture.
Springer –Verlag, 126 -128, June 2009.
Nikola .K. Kasabov, (1998). Foundations of Neural Network,
Fuzzy Systems, and Knowledge Engineering. A
Bradford Book, The MIT press, Cambridge,
Massachusetts, London. Copyright Massachusetts
Institute of Technology.
Otim Nape G. W., Alicai T. and Thresh J.M., (2001).
Changes in the incidence and severity of Cassava mosaic
virus disease, varietal diversity and Cassava Production
in Uganda. Association of applied Biologists.
Ozgur Kisi.(2009). Modelling monthly evaporation using
two different neural computing techniques. Springer
–Verlag.
Pallavi B.E (2005): A Thesis titled "Computer-Aided
Diagnosis Tool for the Detection of Cancerous Nodules
in X-ray images" submitted to Andhra University
Visakhapatnam, India.
Reza Afkhami and Mosalman F., (2006). Application of
Neural Networks for Short-Term Load Forecasting.
IEEE.
Seetha. M., Muralikrishna. I.V, Deekshatulu. B.L,
Malleswari. B.L., Nagaratna, P.Hegde P., (2008).
Artificial Neural Networks and other methods of image
classification. Journal of Theoretical and Applied
Information Technology (JATIT).
Senthil Kumar (2013): “Diagnosis of heart disease using
Advanced Fuzzy resolution Mechanism” International
Journal of Science and Applied Information Technology
(IJSAIT), Vol.2 , No.2, Pages : 22-30 (2013).
Sivanandam et.al., (2007): “Introduction to Fuzzy Logic
using Matlab”: Springer Berlin.
Sseruwagi P., Sserubombwe W.S., Legg J.P, Ndunguru J.,
J.M. Thresh J.M., (2004). Methods of Surveying the
incidence and Severity of Cassava Mosaic Disease and
White Fly Vector Populations on Cassava in Africa: a
review. Available online at www.sciencedirect.com,
2004.
Wai –Kai Chen and Dieter .A. Mlynski. (2007). Principles of
Artificial Neural Networks. Daniel Graupe, 2nd edition,
volume 6. World scientific publishing Co Ltd
Wang M.X., Liu G.D, W.L. Wu, Y.H. Bao and W.N. Liu.
(2006). Prediction of Agriculture derived ground water
nitrate distribution in North China plain with GIS –based
BPNN. Springer –verlag.
Wei-Ta Fang, Hone –Jay Chu, Bai-You Cheng. 2009.
Modelling waterbird diversity in irrigation ponds of
Taoyuan, Taiwan using an artificial neural network
approach. Springer –Verlag.
Zweig M.H., Campbelll G., (1993) Receiver-operating
characteristic (ROC) plots: a fundamental evaluation
tool in clinical medicine. Clinical Chemistry, 39,
561-577.
8