Data mining techniques are used to discover the interesting patterns with least amount of time. We have automated data collection tools to Store large amount of data automatically without manual entry and web computerized society.Inside this paper we are discussed about association: Based on conditions we are going to find out the interesting patterns. This we will represented in two parts one is antecedent it will carry out the condition and another part is consequent that contains the action what we have to do if the antecedent part is correct. We narrated about classification that is allocating items in group of data to aimed class. The main aim of this algorithm is to finding the correct class with full accuracy.
in this we have two types of data those are trading data and testing data. training data is utilized to construct or build a model and testing data is to validate the constructed model.in this we are having algorithms like navy bays , neural network-nearest neighbor ,support vector machine algorithm.Minnesota Intrusion prevention System is a kind of data mining method which we have been used to note attacks over computer Networking Systems. There are part of traditional techniques or methods which are stationed consequent to disburse knowledge of attack signatures which are provided by human experts. There is a significant control in signature based technique is that it cannot detect novel attacks. There is a substantial latency if once a new attack is discovered and its signature is developed. Based on data mining it led to an increasing interest in intrusion detection techniques.
Misuse detection and anomaly detection these are the two categories fall under this category. Misuse detection models have large degree of accuracy in finding out the attacks and its differences and their drawback is the inability to find attacks whose instances have yet not been observed. The benefit of anomaly algorithm is their ability to potentially find unforeseen attacks.
Data mining normally refers to the operation of extracting or mining knowledge from a huge amount of data. This process primarily understand the existing data and it will call the new data. Predictive and descriptive are the two categories divided based upon the data mining tasks. The general properties of data in data mining is divided based upon its descriptive mining. Anomaly detection normally proposed two data mining approaches: association rules and frequency episodes. In Association Rule Algorithms correlations are found between feature and attributes that are used to describe a data set.
In association with the proliferation of cyber security threats like malicious viruses and worms, denial of service (Dos) attacks, Online Internet fraud, achieving valuable network intrusion security is very difficult in protecting our information infrastructure. Classification algorithms have been used in network detection. In that neural network it will separate the networks intrusion and as usual working. Support vector machine also do the same thing but apart from that it will extract the needed features of intrusions in intranet. Some authors find out tree and array SVM algorithms in order to overcome the problem of sequential minimum optimization method which is already used in intrusion detection.
K-means approach is related to k-mean algorithm of clustering, that one is renamed as y-means in intrusion detection.