Intrusion detection has become an essential element of network administration thanks to the huge range of attacks persistently threaten our computers. Ancient intrusion detection systems area unit restricted and do not give a whole resolution for the matter. They look for potential malicious activities on network traffics; they generally succeed to search out true security attacks and anomalies. However, in several cases, they fail to observe malicious behaviors (false negative) or they fireplace alarms once nothing wrong within the network (false positive). Additionally, they need thorough manual process and human professional interference. Applying data processing (DM) techniques on network traffic information may be a promising resolution that helps develop higher intrusion detection systems. Moreover, Network Behavior Analysis (NBA) is additionally associate degree effective approach for intrusion detection. During this paper, we tend to discuss DM and NBA approaches for network intrusion observation and recommend that a mix of each approach has the potential to detect intrusions in networks additional effectively.