Razie Hyria

December 4th, 2018

Annotated Bibliography
To assist in the discussion and argument formation within answering the research question of:
“How can we advance artificial intelligence and neural networking, and thus its reliability, as it
is being integrated into more modern age technology?” I’ve composed this annotated
bibliography.

Galily, M., Roudsari, F., & Sadri, M. (2005). Application of computational intelligence (fuzzy logic,
neural networks and evolutionary programming) to active networking technology. Fuzzy
Systems and Knowledge Discovery, Pt 1, Proceedings, 3613, 900-909. (Link).
Researchers Galily, Roudsari and Sadri argue that “The introduction of active
networking adds a high degree of flexibility in customizing the network infrastructure
and introduces new functionality.”(Pg. 900) Thus, the need for investigating both the
applications of computational intelligence (fuzzy logic, neural networks, and
evolutionary programming) techniques in a newer networking environment, as well as the
provisions of active networking technology that can be exploited for improved operation
(Pg. 900). The researchers offer different applications of computational intelligence
techniques in active networking technology, such as within “Routing”(Pg. 909). .

Hager, G. D., Bryant, R., Horvitz, E., Mataric, M., & Honavar, V. (2017). Advances in artificial
intelligence require progress across all of computer science.. (Link).
Hager, Bryant, Horvitz, Mataric, and Honavar team up to review several
promising areas of interaction between AI and broader computer science that
provide rich opportunities ahead for research and development. The researches
exam the intersections of algorithms, computing systems, theoretical computer science,
cyber security, programming languages, human-computer interaction, and more (Pg. 2).
They go on to claim “there is a growing and compelling imperative to leverage the
advances in AI and automation to improve human lives in many ways”(Pg. 7). In looking
at computing systems, the researchers stressed the importance of improving and proper
integration AI with the example: “current uses of cloud computing are typically not hard
real-time. Whether a search query is answered in 0.2s or 0.4s may not matter much in
practice. However, a query from a car moving at 60mph may need to be answered within
a hard real-time constraint to ensure the safe operation of the automobile.” (Pg. 3)

Marks, S. (2017). Immersive visualization of 3-dimensional spiking neural networks. Evolving

Systems, 8(3), 193-201. doi:10.1007/s12530-016-9170-8. (Link).
Marks explains the recent increase and development of artificial neural networks
which has also lead to an increase of performance, but raises the issues that come along
 with it; complexity and size. Marks poses a solution to the challenges of exploration and
analysis of spatial structures and temporal behavior of such networks by presenting:
“NeuVis, a framework for an intuitive and immersive 3D visualization of spiking neural
networks in virtual reality, allowing for a larger variety of input and output devices.” (Pg.
195) Applying NeuVis technology to the “NeuCube, a 3-dimensional spiking neural
networking learning framework” (similar to that of the Oculus Rift) (Pg. 198),
significantly improving the users abilities to explore, and analyze while simultaneously
debugging the network. This approach relies on observations and qualitative
data of around 15 specialists of KEDRI and about 30 non-specialists (Pg. 200). Marks
discusses further venues of development and alternate rendering methods which
ultimately increase the visual accuracy and realism of visualization as well as further
extends its analysis and exploration capabilities.

Parekh, R., Yang, J., & Honavar, V. (2000). Constructive neural-network learning algorithms for
pattern classification. IEEE Transactions on Neural Networks, 11(2), 436-451.
doi:10.1109/72.839013. (link)
Mathematicians Parkh, Yang, and Honavar analyze constructive learning
algorithms as an approach for the incremental construction of near-minimal neural-networking
architectures for pattern classification. Said algorithms
are proposed in the literature and shown to converge to zero classification errors. The
mathematicians present “two constructive learning algorithms MPyramid-real and MTiling-real
that extend the pyramid and tiling algorithms, respectively, for learning real to M-ary mappings
(i.e., classification problems involving real-valued input attributes and multiple output classes)”
(Pg. 436). Parkh, Yang, and Honavar work with an extensive amount of mathematical proofs
such KJCjP+kOqL-1j where k+p-l=kj, i.e kj< k (Pg. 441) to prove the convergence of these
algorithms and empirically demonstrate their applicability to practical pattern classification
problems, ultimately concluding which is more efficient with the progressing technology
(Pg.449).

Vourvopoulos, A., Bermudez i Badia, S., & Liarokapis, F. (2017). EEG correlates of video game
experience and user profile in motor-imagery-based brain–computer interaction. The
Visual Computer, 33(4), 533-546. doi:10.1007/s00371-016-1304-2. (link)
Vourvopoulos, Badia, and Liarokapis argue that experienced gamers could have
better performance in BCI training due to enhanced sensorimotor learning derived from
gaming (Pg. 533). The researchers were motivated by the advancement of immersive
virtual environments and 3D worlds of neurogames. Their argument is investigated by
“two experimental studies with 20 participants overall, undergoing 3 BCI sessions,
resulting in 88 EEG datasets” (Pg. 536). The research is focused on the effect that gaming
experience has in brain activity modulation as an attempt to be utilized in neurogame
design (Pg. 533).