![](bg2.png)
AI has two the main purposes: creating expert
systems and implement human intelligence in
machines. Relatively to the first one, such systems
should exhibit intelligent behavior and the ability to
learn, demonstrate, explain, and advise its users. The
second one involves the creation of systems that
understand, think, learn, and behave like humans
(Nilsson, 2014). In addition, it is important to realize
that AI is present in very different areas with very
different goals and approaches. For example, AI plays
a crucial role in strategic games such as chess, poker,
etc., where the machine has the capability of thinking
in a large number of possible positions based on
heuristic knowledge. In addition to this, certain
intelligent systems are capable of hearing and
understand the language in terms of sentences and
their meanings while a human talk to those (Nilsson,
2014).
At the end, it can be concluded that AI can have
the following functions: solving problems, pattern
recognition, classification, learning, induction,
deduction, building analogies, optimization,
surviving in an environment, language processing,
knowledge and much more (Silva, 2016, Hutter,
2005).
2.2 Machine Learning
Machine learning is one subfield of Artificial
Intelligence. It hugely relies on mathematical
algorithms that improve learning through experience,
in an attempt to build systems that learn from past
data, conceding predictions and recognition of
patterns (Hutter, 2005, Hamet, 2017).
There are three types of machine learning
algorithms:
Unsupervised: the capability of finding
patterns.
Supervised: algorithms of classification and
regression that have in consideration the
previous data.
Reinforcement learning: use of sequences of
rewards and punishments to create a strategy
to operate in a certain problem space
(Hamet, 2017).
Machine learning is growing through the time,
bringing new discoveries and utilities into different
areas such as science and engineering. For instance,
by resorting to machine learning techniques, now it is
possible to measure the detailed molecular state of an
organism.
In that way, the main goal of machine learning is
to infer a functional relationship between a set of
attributes variables and associated response or target
variables in order to predict the response for any set
of attributes, where such response can be the result of
classification, clustering or projection (Rogers,
2017).
2.3 Data Mining
Data Mining is a process that tries to discover
unknown, unexpected, interesting relevant patterns
and relationships in data that may be used to make
valid and accurate predictions, by using a large
variety of data analysis methods (Stubee, 2014). Data
Mining involves techniques from different disciplines
such as database technology, statistics, machine
learning, high-performance computing, pattern
recognition, neural networks, data visualization,
information retrieval, image and signal processing,
and spatial data analysis (Han, 2000, Esteves, 2017).
This process tries to achieve useful knowledge
through huge amounts of data, in a process of
extraction of new information. Such knowledge can
be used in applications of business management,
production control, market analysis, engineering
design, medicine, science exploration, etc. (Han,
2000).
Data Mining is the analysis of data sets aiming to
find unsuspected relationships and to summarize the
data in new ways that are understandable and useful
to the data owner. For that, several methods are used,
including linear equations, rules, clusters, graphs, tree
structures, and recurrent patterns in time series (Hand,
2010).
Such processes require the use of large datasets. If
only small data sets were used, we would obtain
instead a classical discussion of exploratory data
analysis, practiced by statisticians (Hand, 2010).
In that way, it can be concluded that Data Mining
is used to identify potential problems and to discover
similarities between current and previous situations,
in order to improve the understanding of relevant
factors and associations as well as discovering non-
obvious features in the data (Cortès, 2000).
Data Mining is used in different areas in order to
make easier discovering unknown and significant
information to an organization. Following there are
some examples of research works, which used such
process of learning.
The paper "Real-time Decision Support using
Data Mining to Predict Blood Pressure Critical
Events in Intensive Medicine Patients" has the main
purpose of predicting the probability of a patient
having a blood pressure critical event by using Data
Mining classification techniques (Portela, 2015).
HSP 2018 - Special Session on Healthy and Secure People
286