recognition is, what we see today in smartphones, the technology that allows
users to issue commands using their hand movements. It is used for smartphones’
extra features, gaming, and sign language interpretation.
an academic article from September 2017 by KSII Transactions on Internet and
Information Systems, a gesture recognition system is being developed based on
orientation, linear extrapolation and the longest common subsequence algorithm.
recognition systems uses motion and skin color, measured by a 3-frame
difference and the skin color elliptic boundary model, to distinguish a hand in
convoluted environment. The 3-frame reference involves using a video camera to
detect motion by turning 3 consecutive frames into a grayscale intensity images
where the skin-color elliptic boundary model differentiates the hand from other
objects and gauges the distance and velocity of the hand.
algorithms for the recognition process are based on the algorithm, called the
longest common subsequence. The longest common subsequence is an algorithm
based on recursion (which we covered in the course) which goes through a
sequence of data and derives a subsequence, and repeats said process. This
algorithm is used to compare the inputted hand gestures performed and analyze
which is the most probable and applicable gesture pattern on the already
explored article, a suggestion for improvement in the gesture recognition
system is considering the obstruction or overlapping of hands, the orientation of
the hands and their velocity. From this information, one might be able to
create an algorithm to predict the linear motion of the hand so that the
technology can track the hands.
et al. “A Dynamic Hand Gesture Recognition System Incorporating
Orientation-based Linear Extrapolation Predictor and Velocity-assisted Longest
Common Subsequence Algorithm.” KSII Transactions on Internet and
Information Systems, vol. 11, no. 9, 2017, p. 4491+. Academic OneFile,
Accessed 13 Jan. 2018.
is a computer science theoretical field where patterns are sought after in
large datasets. I found an article discussing how could one use data mining in
a non-linear dataset. The problem lies within the inputs and the subsets of the
dataset, as they grow exponentially. Thus, it makes it more difficult to set
parameters and transformations that will cover the predicted data. Assuming enough
information and subsets are given to create the correct parameters, one would
use a forward selection method or a backward elimination technique. The forward
selection method goes through each subset and adds the ones with significant effects
on the results as deemed by the analyst’s (of said data) criteria of
significance as predictors. While the backwards technique starts with all the
predictors and disposes of the subsets that would not change the results of the
dependent variable significantly.
there is a method of the correct subsets, called Least Angle Regression. This
method is based on the front method mentioned previously. In this method, the coefficients
of the independent or inputted values are to be zero. First predictor changes its coefficient to
its least square value, and so is done to the rest of the subsets, where if one
subset matches the first, they are added to the set.
transformations, a recursive Newton algorithm is incorporated, as well as a
machine learning method, called Artificial Neural Networks.
Peyman, et al. “A non-linear data mining parameter selection algorithm for
continuous variables.” PLoS ONE, vol. 12, no. 11, 2017, p. e0187676.
Accessed 13 Jan. 2018.
medical field, there are large datasets of information, ontologies, and medical
terms to be sorted, arranged and cross referenced. The branch of computer science
that creates programs to make sense of biological data is called bioinformatics,
and intersects with the medical field. So, in the bioinformatics, access to the
vast scientific information that comes in all its different forms is crucial
for studies, research and breakthroughs.
BioMed Central Ltd., the Research Centre is working on a web-based platform,
called PIBAS FedSPARQL where researchers can access information. They
implemented a text mining algorithm to allow users to query a given data item,
but to do this, they set up predefined SPARQL endpoints holding different types
of datasets, where similar data items would be found in the SPARQL endpoints.
data items and searching for similar ones is one of the issues faced by those
in bioinformatics. In FedSPARQL, one way to find similarity is by calculating
the shortest distance between two words (concepts) in a taxonomy of concepts.
to this, since science is a progressing field and research is done daily, new
information would be added to the database. Thus, there is a predefined template
for users to add their datasets with their finding. The issue with this would
be that the small laboratories would not be contributed or harder to regulate.
significantly is similar to the course material we covered in ICS4U, as we
learned how handle datasets and manipulate them, as well as search information
and sort according to categories, so do people who work in bioinformatics.
Marija, et al. “PIBAS FedSPARQL: a web-based platform for integration and
exploration of bioinformatics datasets.” Journal of Biomedical Semantics,
vol. 8, no. Suppl 1, 2017. Academic OneFile,
Accessed 13 Jan. 2018.
Entering and breaking through the firewalls of a computer without warrant is an
ethical dilemma to some. One of the reasons can be the intentions behind their
hacking into a system is purely out of curiosity or a thirst of knowledge. They
may not corrupt or copy files, but simply look at them. That can be considered
a “victimless” crime, as no one was hurt in the process. Moreover, some may
argue that not all information stored in a computer is private or sensitive information
in need of protection. Nevertheless, despite the lack of physical harm, a
breach of privacy had occurred, similar to breaking and entering into a house.
In addition to this, the simple exposure of information may go against the
confidentiality of some companies.
there are university professors who assign their students the task of hacking
into the university’s system, but is this seen as a breach of privacy, as the
information is not clearly stated to which individual it belongs to. As well,
the perpetrators had permission from a figure of authority, but is it the
correct figure of authority.
the Institute of Electrical and Electronics Engineers, there are 8 different
moral principles for software engineers to comply with, some of which are:
1. Public – software
engineers should act and take into consideration of desires of the public.
Software engineers should not false advertise their product, much less
incorporate insensitive/offensive material into their programs. Moreover, creators
should give fair warnings upon the use of their product, for instance medical
2. Client and Employer
– they should act in accordance of the client’s and the public’s interest
Software engineers must not accept payments from competitive companies to
sabotage their product. They must report on their progress and express their truthful
opinion to their clients in terms of budget, schedule, or the rate of success
of the project.
Sodtware engineers must produce high-quality work
should not be a copied software, only original work. They must only make the product
or accept the responsibility to make it if they are capable of performing the
task. They must persevere to meet all the requirements and specifications set
by the client in regards to the projects.
4. Profession –
they should act in a respectful manner to maintain the reputation of the field.
Moreover, they must ensure to never place their interests above their clients’ or
A code like
this needs to exist, as computers introduced a whole new plane of moral dilemmas
and uncertainties. Crimes can now be committed in way never done before. In
order to protect human values, one needs to establish a code for computer
programmers to provide clarification.