Page 172 - DCAP506_ARTIFICIAL_INTELLIGENCE

Artificial Intelligence

Notes 14. A ......................... is nothing but a great production that does the work of the whole sequence
of smaller ones.

12.9 Discovery-based Learning

Discovery is a limited form of learning. The knowledge acquisition is performed without
obtaining any support from an educator. Discovery Learning is an inquiry-dependent learning
method.

In discovery learning, the learner utilizes his own experience and prior knowledge to learn the
truths that are to be studied. The learner builds his own knowledge by researching with a
domain, and inferring rules from the outcomes of these experiments. In addition to domain
information the learner require some support in selecting and interpreting the information to
construct his knowledge base.
A cluster is a compilation of objects which are alike in some manner. Clustering groups data
items into resemblance classes. The properties of these classes can then be utilized to understand
problem traits or to discover similar groups of data items. Clustering can be defined as the
process of decreasing a huge set of unlabeled data to controllable piles involving similar items.
The similarity measures rely on the assumptions and preferred handling one brings to the data.
Clustering starts by doing trait withdrawal on data items and compute the values of the selected
feature set. Then the clustering model chooses and compares two sets of data items and outputs
the similarity measure among them. Clustering algorithms that utilize specific similarity
measures as subroutines are engaged to construct clusters.
The clustering algorithms are usually classified as Exclusive Clustering, Overlapping Clustering,
Hierarchical Clustering and Probabilistic Clustering. The collection of clustering algorithms
depends on different criteria like time and space complexity. The outcomes are verified to
observe if they meet the standard or else some or all of the above steps have to be frequent.

Some applications of clustering are data compression, hypothesis generation and hypothesis
testing. The theoretical clustering system accepts a set of object descriptions in the form of
events, observations, facts and then generates a classification method over the observations.

Notes COBWEB is an incremental abstract clustering system. It incrementally adds the
objects into a categorization tree. The striking trait of incremental systems is that the
knowledge is modernized with every new observation. In COBWEB system, learning is
incremental and the knowledge it learned in the form of classification trees raise the
inference capabilities.

Task Illustrate the concept of clustering system.

Self Assessment

Fill in the blanks:

15. In ......................... learning, the learner utilizes his own experience and prior knowledge to
learn the truths that are to be studied.

166 LOVELY PROFESSIONAL UNIVERSITY