Intrusion Detection and Forensics based on decision tree and Association rule mining for Probe attack detection

International
OPEN ACCESS Journal
Of Modern Engineering Research (IJMER)
| IJMER | ISSN: 2249–6645 | www.ijmer.com | Vol. 5 | Iss.4| Apr. 2015 | 31|
Intrusion Detection and Forensics based on decision tree and
Association rule mining for Probe attack detection
Harishchandra Maurya1
, Swati Sharma2
1
M .Tech (C.S) (Assistant Professor, Bhagwant University, Ajmer, India)
2
M .Tech (C.S) (Scholar, Bhagwant University, Ajmer, India)
Department of Computer Sciences and Engineering Bhagwant University, Ajmer, India
I. Introduction
Intrusion detection [1] is becoming an increasingly important technology that monitors network traffic
and identifies network intrusions such as anomalous network behaviours, unauthorized network access, and
malicious attacks to computer systems. There are two general categories of intrusion detection systems (IDSs):
misuse detection and anomaly detection. Misuse detection systems detect intruders with known patterns, and
anomaly detection systems identify deviations from normal network behaviours and alert for potential unknown
attacks. In past years, there were only few intruders and so the user could manage them easily from the known
or unknown attacks, but in recent years the security is the most serious problem. Because the intruders introduce
a new variety of intrusions in the market, so that the user can’t manage the computer systems and networks
properly. Different approaches and algorithms are used to detect the attack. In this paper we propose the
techniques which can detect network based attacks using Association rule mining [2] and decision tree [3].
Computer forensics [4]. has significant ability to make network infrastructures more integrated and
capable of surviving attack. Computer Forensics is the use of computer technology, in accordance with the
lawful procedures and criterion. From the evidence points to technical analysis, computer forensics technology
mainly divided into static and dynamic forensics evidence. Dynamic forensics technology will be integrated into
the firewall, intrusion detection. Dynamic computer forensics combined with the intrusion Detection, which can
collect reliable evidence in real time when the system is invaded, complete the invasion of Testing and the
evidence of the dynamics of computer forensics, becomes research focus of computer forensics.
 Decision tree
Decision trees are well known machine learning techniques. A decision tree is composed of three basic
Elements.
 A decision node specifying a test attributes.
 An edge or a branch corresponding to the one of the possible attribute values this means one of the test
attribute outcomes.
 A leaf which is also named an answer node, contains the class to which the object belongs.
In decision trees, two major phases should be ensured:
ABSTRACT—This paper present an approach based on the combination of, two techniques using
decision tree and Association rule mining for Probe attack detection. This approach proves to be
better than the traditional approach of generating rules for fuzzy expert system by clustering methods.
Association rule mining for selecting the best attributes together and decision tree for identifying the
best parameters together to create the rules for fuzzy expert system. After that rules for fuzzy expert
system are generated using association rule mining and decision trees. Decision trees is generated for
dataset and to find the basic parameters for creating the membership functions of fuzzy inference
system. Membership functions are generated for the probe attack. Based on these rules we have
created the fuzzy inference system that is used as an input to neuro-fuzzy system. Fuzzy inference
system is loaded to neuro-fuzzy toolbox as an input and the final ANFIS structure is generated for
outcome of neuro-fuzzy approach. The experiments and evaluations of the proposed method were
done with NSL-KDD intrusion detection dataset. As the experimental results, the proposed approach
based on the combination of, two techniques using decision tree and Association rule mining
efficiently detected probe attacks. Experimental results shows better results for detecting intrusions as
compared to others existing methods.
Keywords—Intrusion detection, Forensics, decision tree, Association rule mining, Probe attack.

Intrusion Detection and Forensics based on decision tree and Association rule mining for…
Building the tree: Based on a given training set, is to build a decision tree. It consists of each decision node,
Select the "appropriate" test properties, and define the class label of each leaf.
Classification: Order to classify a new instance; we began to determine the root of the tree, then we test the
Node specified property. The test results, allowing moving down the tree relative to a given instance of the
attribute value. This process is repeated until it encounters a leaf. The instance is then classified in the same
class Characteristics to leaves.
Decision tree as a predictive model which maps observations about an item to conclusions about the
item's target value. More descriptive names for such tree models are classification trees or regression trees. In
these tree structures, leaves represent class labels and branches represent conjunctions of features that lead to
those class labels. In decision analysis, a decision tree can be used to visually and explicitly represent decisions
and decision making. In data mining, a decision tree describes data but not decisions; rather the resulting
classification tree can be an input for decision making. Data comes in records of the form:
(x,Y) = (x1, x2, x3, x4,…, xk, Y) (1)
The dependent variable, Y, is the target variable that we are trying to understand, classify or generalize. The
vector x is composed of the input variables, x1, x2, x3 etc., that are used for that task.
Figure 1: General structure of decision tree
 Association rule mining
Association rule mining is one of the hottest research areas that investigate the automatic extraction of
previously unknown patterns or rules from large amounts of data. The basic concept of association rule mining is
given below:
Let I = {I1, I2, …, Im} be a set of items. Let D be a database of transactions where each transaction T is a set of
items such that T I. Each transaction is associated to an identifier, call TID. A transaction T is said to contain A
if and only if A T. An association rule is an implication of the form A B, where A I, B I, and A B = . The rule A
B holds in the transaction set D with support s, where s is the percentage of transactions in D that contain AB.
The rule AB has confidence c in the transaction set D. That is,
(1)
(2)
Where A is named as the support count of the set of items A in the set of transactions D, as denoted by
sup_count(A). A occurs in a transaction T, if and only if AT. Rules that satisfy both a minimum support threshold
(min_ sup) and a minimum confidence threshold (min_conf) are called strong. A set of items referred to as an
itemset. An itemset that contains k items is a k -itemset. Itemsets that satisfy min_ sup is named as frequent
itemsets. All strong association rules result from frequent item sets.
 Probing Attack and Its Type
Probing Attack [5]. is an attempt to gather information about a network of computers for the apparent purpose of
circumventing its security controls. It involves discovering the algorithms and parameters of the recommender
system itself. It may be necessary for an intruder to acquire this knowledge through interaction with the system
itself.
 Ipsweep.
 Portsweep.
 Nmap
 Satan
The subsequent parts of this paper are organized as follows: At first, in section 2, presents the proposed approach.
Next, at section 3 and 4, the proposed approach procedure of Probe attack detection is explained and experimental
results are discussed respectively. Finally, section 5 makes some concluding remarks and proposes further areas
for research.

II. Proposed Approach
This proposed approach uses combination of, two techniques using decision tree and Association rule
mining for Probe attack detection. Decision tree based technique is applied to find out the best suitable
parameters. Some of the classification algorithm that most commonly used to classify the dataset named J48. It is
a collection of machine learning algorithms for data mining tasks. It contains tools for association rules. Data
mining techniques have been commonly used to extract patterns from sets of data. Specifically, association rules
is one of the data mining approach [6] and used for anomaly detection. Association rule find correlations between
features or attributes used to describe a data set. Here, we are combining both the approaches of association rule
mining for selecting the best attributes together and decision tree for identifying the best parameters together to
create the rules for fuzzy expert system. The combination proves to be good technique. Figure below shows the
proposed framework to generate rules for fuzzy expert system.
Figure 2: Generating rules for fuzzy expert system
III. Procedue Based On Decision Tree And Association Rule Mining
For Probe Attack Detection
In this section, we explain the procedure of Probe attack detection. Procedure is as follows:
 In our approach we are proposing a combination of neural and fuzzy systems are two standalone
techniques to efficiently detect the anomalous behaviour. In which fuzzy systems provide a powerful
framework for expert knowledge representation, while neural networks provide learning capabilities and
exceptional suitability for computationally efficient hardware implementations.
 Generates fuzzy rules from the uploaded dataset using subtractive clustering method. This approach of
generating rules using clustering methods is limited to the problems of clustering techniques. The NSL-
KDD dataset [7]. is not suitable for clustering [8]. So we are purposing the new method to generate rules
for fuzzy expert system using association rule mining and decision trees.
 Our work is implemented using MATLAB 7.6.0 (R2008a) Neuro-Fuzzy Toolbox software. MATLAB
provides several computing tools based GUIs to work with. In this dissertation work we are using Neuro-
Fuzzy toolbox (i.e. anfisedit) and Fuzzy Logic toolbox[9] (i.e. fuzzy) to implement our work
 The NSL-KDD is used as database to train and test the system performance. Experiments are performed
with NSL-KDD dataset. We have conducted experiments on Probe attack.
 Both training and testing dataset can loaded in ANFIS tool.
 Decision Trees in WEKA tool is used to analyze the NSL-KDD and identify the required parameters.
 After that Membership functions are generated using trapezoidal functions and one Gaussian function is
used for Boolean type of attribute (i.e. logged_in). Nine inputs having three membership functions except
loggen_in (i.e. having 2 membership functions) generated for the probe attack. Figure 3.1 shows the
generated fuzzy membership functions and their attributes.

Figure 3.1: Membership functions for Probe Attack
 Based on these rules we have created the fuzzy inference system that is used as an input to neuro-
fuzzy system. Figure 3.2 shows the generated FIS [10].
Figure 3.2: Fuzzy inference system
 Figure 3.3 shows the output of the fuzzy inference system. This system can be adjusted according to
the values of the input attributes. Fuzzy Inference system shows an accuracy of 94.4 %.
Figure 3.3: Fuzzy inference system output

 this fuzzy inference system is loaded to neuro-fuzzy toolbox as an input and the final ANFIS[11]
structure is generated as shown in figure 3.4 below
Figure 3.4: ANFIS structure
 ANFIS tool output we have implemented it for detecting probe attacks. We have implemented it for
detecting probe attacks. Neuro-Fuzzy system results in 99.68 % accuracy in detecting the probe
attack.
IV. Experiment And Results
In this section, we summarize our experimental results to detect Probe attack using decision tree and
Association rule mining over NSL-KDD dataset. We have conducted experiments on Probe attack having nine
contributing attributes having 1000 number of instances.NSL-KDD dataset consists of 1000 instances in training
dataset and similar instances in testing dataset that can be loaded easily into the neuro-fuzzy toolbox. The attack
types are grouped into four categories:
 Ipsweep: It probes the network to discover available services on the network. First intruder find out a
machine on which he may be attacked.
 Portsweep: It probes a host to find available services on that host. If a service is known on the system so
it may easily be attacked by the network intruder.
 Nmap: It is a complete and flexible tool for scanning a network either randomly or sequentially.
Therefore, often intruders used this tool for scanning network parameters that may help them in attacking
the system.
 Satan: It is an administration tool; it gathers information about the network. This information can be used
by an attacker.
TABLE1: TRAINING DATASET REPRESENTING PROBE ATTACK ATTRIBUTES
1 2 3 4 5 6 7 8 9
src_by
tes
logged
_in
serror_
rate
srv_serror
_rate
rerror_
rate
same_srv_
rate
diff_srv_
rate
dst_host_srv_
count
dst_host_serro
r_rate
491 0 0 0 0 1 0 25 0
146 0 0 0 0 0.08 0.15 1 0
0 0 1 1 0 0.05 0.07 26 1
232 1 0.2 0.2 0 1 0 255 0.03
199 1 0 0 0 1 0 255 0
0 0 0 0 1 0.16 0.06 19 0
0 0 1 1 0 0.05 0.06 9 1
0 0 1 1 0 0.14 0.06 15 1
0 0 1 1 0 0.09 0.05 23 1

0 0 1 1 0 0.06 0.06 13 1
0 0 0 0 1 0.06 0.06 12 0
0 0 1 1 0 0.02 0.06 13 1
287 1 0 0 0 1 0 219 0
0 0 1 1 0 0.17 0.05 2 1
TABLE 2: TESTING DATASET REPRESENTING PROBE ATTACK ATTRIBUTES
1 2 3 4 5 6 7 8 9
src_byte
s
logged_in
serror_ra
te
srv_serror_r
ate
rerror_rat
e
same_srv_r
ate
diff_srv_rat
e
dst_host_srv_c
ount
dst_host_serror
_rate
44 0 0 0 0 0.75 0.5 254 0
0 0 1 1 0 1 0 79 0.21
53 0 0 0 0 1 0 255 0
0 0 1 1 0 1 0 1 1
54540 1 0 0 0.5 1 0 229 0
0 0 0 0 1 0.04 0.06 9 0
0 0 0 0 0 1 0 165 0.49
56 0 0 0 0 1 0 255 0
192 0 0 0 0 1 0 51 0
0 0 0.14 0 0.86 0 1 1 0.13
0 0 0.07 0 0.93 0 1 1 0.07
21 0 0 0.21 0.75 0.5 0.75 40 0
This dataset (i.e. shown in table 1 and 2) is normalized to the range of [0-1]. Standard mean deviation is
used to define the mid value for each rule. The testing dataset, against which the ANFIS system is validated. The
dataset for the Probe attack having 1000 number of instances is given as an input to the ANFIS tool. Both training
and testing dataset can be loaded at the same time while testing dataset is used later to verify the training dataset
inputs. Before creating the fuzzy membership functions, the required parameters range of parameters needs to be
identified. Decision Trees is used to identify the required parameters. The minimum and maximum is set to the
range 0-1. Each attribute is distributed over a range of low, mid (medium) and high. Range of parameters is
shown in table 3.
TABLE 3: MEMBERSHIP PARAMETERS FOR PROBE ATTACK
Attributes Min Low Mid High Max
source_bytes 0 0.2 0.34 0.7 1
logged_in 0 - - - 1
serror_rate 0 0.295 0.453 0.89 1
srv_serror_rate 0 0.294 0.454 0.81 1
rerror_rate 0 0.131 0.335 0.682 1
same_srv_rate 0 0.126 0.44 0.643 1
diff_srv_rate 0 0.069 0.483 0.732 1
dst_host_srv_rate 0 0 0.426 0.436 1
dst_host_serror_rate 0 0.296 0.452 0.691 1
Based on these rules we have created the fuzzy inference system. These rules are used to makes decision
for neuro-fuzzy system. The final outcome is the production of these rules. Fuzzy inference system can be
adjusted according to the values of the input attributes. Fuzzy inference system is used as an input to neuro-fuzzy
system and shows an accuracy of 94.4 %. The last attribute shows the defuzzified output of the fuzzy system.
Fuzzy inference system is loaded to neuro-fuzzy toolbox as an input and the final ANFIS structure is generated.
ANFIS module is trained using neural network back-propagation algorithm. While training the dataset

with this ANFIS structure it is observed that training error reduces with number of epochs. For our experiment we
have set number of epochs to 100. Neural network learn through this phase while reduces the network errors.
ANFIS tool output is evaluated on the basis of accuracy in predicting the attacks clearly. We have implemented it
for detecting probe attacks. Results below show the outcome of neuro-fuzzy approach. The output of neuro-fuzzy
toolbox is the weighted neurons that are defuzzified using centroid defuzzification method. Euro-Fuzzy system
results in 99.68 % accuracy in detecting the probe attack. A slight comparison to previous literature is given as
follows:
TABLE 4: COMPARISON OF RESULTS
Approach in [paper] Probe Detection rate (%)
[84] 73.20
[85] 100
[86] 72.8
[87] 99
ANFIS Approach 99.68
Figure 4.1: Comparison of results
Figure 4.2: Plotting error against number of epochs
Figure 4.2 shows that error decreases with the number of epochs. Error is reduced to 0.58729 to 0.078525 in 100
epochs (iterated seven times). To detect Probe attack using decision tree and Association rule mining is
implemented using MATLAB [12] 7.6.0 (R2008a) Neuro-Fuzzy Toolbox software.
V. Conclusion
In this paper, an approach based on the combination of, two techniques using decision tree and
Association rule mining for Probe attack detection for Intrusion Detection and Forensics was successfully
demonstrated its usefulness on the training and testing subset of NSL-KDD dataset. Well known machine
learning technique called decision tree technique is applied to the NSL-KDD dataset to find out the best suitable
parameters. We used WEKA tool to generate decision trees for analysing NSL-KDD dataset. One of the data
mining approach called association rules is used for selecting the best attributes. Probe attacks are the kind of
attack which tries to collect the data and find the network’s vulnerability. Probing attacks can steel important
information from computer or network and later may be used by the intruder. These attacks may also result in
significant loss of time and money for many organizations. There are many methods for their prevention but has
some limitations like varying nature of attacks. At last, this paper proposed an approach based on a combination
of, two techniques decision tree and Association rule mining. This approach is the combination of these two
technique is use to detect Probe attack category. Experimentation results showed the proposed method is
effective in Probe attack detection. In the future, we will study more on improving detection accuracy of probe
attacks. In addition, we will only use the most relevant features on the given probe detection task to reduce the
computation time and hopefully to enhance the accuracy of detection as well.
REFERENCES
[1] B. Mukherjee, L. T. Heberlein, and K. N. Levitt, "Network intrusion detection," Network, IEEE, vol. 8, no. 3, pp. 26-41,
1994.
[2] J. Luo and S. Bridges, “Mining fuzzy association rules and fuzzy frequency episodes for intrusion detection,” in
International Journal of Intelligent Systems, pp. 687–703, John Wiley & Sons, 2000.
[3] Z. Liu and D. Feng, "Incremental Fuzzy Decision Tree-Based Network Forensic System," in Computational Intelligence
and Security. vol. 3802: Springer Berlin Heidelberg, 2005, pp. 995-1002.
[4] G. Palmer, "A road map for digital forensic research," First digital forensic research workshop (DFRWS'01), pp. 27-30,
2001.
[5] J. McHugh, "Testing Intrusion detection system: a critique of the 1998 and 1999 DARPA intrusion detection system
evaluations as performed by Lincoln Laboratory," ACM Transaction on Information and system security, vol. 3, no. 4, pp.
262-294, 2000.
[6] "WEKA – Data Mining Machine Learning Software," [Available Online]. http://www.cs.waikato.ac.nz/ml/weka/.
[7] M. Sabhnani and G. Serpen, "Application of Machine Learning Algorithms to KDD Intrusion Detection Dataset within
Misuse Detection Context. [Available At] http://www.eecs.utoledo.edu/~serpen/.
[8] M. B. Al-Zoubi, A. Hudaib, and B. Al-Shboul, "A fast fuzzy clustering algorithm," in in Proceedings of the 6th WSEAS Int.
Conf. on Artificial Intelligence, Knowledge Engineering and Data Bases, Corfu Island, Greece, 2007.
[9] Fuzzy Logic Toolbox, [Available At] http://www.mathworks.in/help/fuzzy/index.html
[10] J. S. R. Jang, "ANFIS: adaptive-network-based fuzzy inference system," Systems, Man and Cybernetics, IEEE
Transactions, vol. 23, no. 3, pp. 665-685, 1993.
[11] A. Al-Hmouz, S. Jun, R. Al-Hmouz, and Y. Jun, "Modeling and Simulation of an Adaptive Neuro-Fuzzy Inference System
(ANFIS) for Mobile Learning," IEEE Transactions on Learning Technologies, vol. 5, no. 3, pp. 226-237, 2012.
[12] C. Moler, "The Origins of MATLAB,", Retrieved April 15, 2007, [Available Online]
http://www.mathworks.in/products/matlab/, 2004.

Intrusion Detection and Forensics based on decision tree and Association rule mining for Probe attack detection

More Related Content

What's hot (14)

Viewers also liked (20)

Similar to Intrusion Detection and Forensics based on decision tree and Association rule mining for Probe attack detection (20)

More from IJMER (20)

Recently uploaded (20)

Intrusion Detection and Forensics based on decision tree and Association rule mining for Probe attack detection