Page 1442 - (ISC)² CISSP Certified Information Systems Security Professional Official Study Guide
P. 1442

updating models of activity.

               Machine learning techniques fall into two major categories.


                    Supervised learning techniques use labeled data for training. The
                    analyst creating a machine learning model provides a dataset along
                    with the correct answers and allows the algorithm to develop a
                    model that may then be applied to future cases. For example, if an
                    analyst would like to develop a model of malicious system logins,
                    the analyst would provide a dataset containing information about
                    logins to the system over a period of time and indicate which were
                    malicious. The algorithm would use this information to develop a

                    model of malicious logins.

                    Unsupervised learning techniques use unlabeled data for training.
                    The dataset provided to the algorithm does not contain the
                    “correct” answers; instead, the algorithm is asked to develop a
                    model independently. In the case of logins, the algorithm might be
                    asked to identify groups of similar logins. An analyst could then

                    look at the groups developed by the algorithm and attempt to
                    identify groups that may be malicious.


               Neural Networks

               In neural networks, chains of computational units are used in an
               attempt to imitate the biological reasoning process of the human
               mind. In an expert system, a series of rules is stored in a knowledge

               base, whereas in a neural network, a long chain of computational
               decisions that feed into each other and eventually sum to produce the
               desired output is set up. Neural networks are an extension of machine
               learning techniques and are also commonly referred to as deep
               learning or cognitive systems.

               Keep in mind that no neural network designed to date comes close to

               having the reasoning power of the human mind. Nevertheless, neural
               networks show great potential to advance the artificial intelligence
               field beyond its current state. Benefits of neural networks include
               linearity, input-output mapping, and adaptivity. These benefits are
               evident in the implementations of neural networks for voice
               recognition, face recognition, weather prediction, and the exploration
   1437   1438   1439   1440   1441   1442   1443   1444   1445   1446   1447