Cyber Security Works
Home   About Us   Technology   Solutions   Research   Company  
Network Penetration Testing
Application Penetration Testing
Digital Forensics
Data Mining
Training & Education
Technical papers in the field of Information Security
FREE Quote & Consultation
Data Mining and Knowledge Recovery

Knowledge discovery and data mining construes of techniques that assists in discovering associations within a given dataset.  A few techniques that facilitate extraction of knowledge from data are machine learning, soft computing, statistical techniques, pattern recognition, and visualization. Knowledge extracted from large datasets can be generalized and further utilized in a variety of similar problems.

Cyber Security Works Pvt. Ltd research scientists have great expertise in web mining, text mining, legal informatics, bioinformatics, clinical informatics, prognosis analysis, social networking, anomaly detection, including online fraud detection, fraudulent financial transaction detection, and intrusion detection (e.g., wireless sensor network anomaly detection). Our researchers developed new feature selection and feature ranking techniques that are widely used in security related applications and bioinformatics.

Algorithm and Model Development

  • Behavior mining
  • Pattern matching
  • Inter-transaction association mining
  • Discovering hidden patterns and relationships

Feature and Model Selection

Heuristics based feature ranking techniques that consider the performance of a particular learning machine for the ranking of features
  • Forward Selection: features are progressively incorporated into larger and larger subsets
  • Backward Elimination: least promising features are progressively eliminated
Embedded methods, feature ranking is incorporated as part of the training process
  • CART (a decision tree methodology)
Learning machine specific feature ranking methods
  • Support vector decision function (SVDF)
  • Multivariate regression spines (MARS)
  • Linear genetic program (LGP)

Anomaly Detection

  • K-nearest neighbor (KNN)
  • Fuzzy c-means
  • Self organizing maps
Instance based learning
  • Case based reasoning
  • K-nearest neighbor (KNN)
  • Radial basis function network (RBF)

Signature Based Detection

  • Support vector machines (SVM)
  • Linear genetic programs (LGP)
  • Artificial neural networks (ANN)
  • Adaptive network based fuzzy inference systems (ANFIS)
  • Classification and regression trees
  • Statistical methods
  • Random forests
  • Kernel based methods


Contact Us
About Us
|  Overview
|  Network Penetration Testing |  Application Penetration Testing
|  Digital Forensics |  Data Mining
|  Training & Education
|  Published Papers