STORRE Collection: Electronic theses of Computing Science and Mathematics students.Electronic theses of Computing Science and Mathematics students.http://hdl.handle.net/1893/362024-03-29T11:35:51Z2024-03-29T11:35:51ZModeling hippocampal theta-coupled gamma oscillations in learning and memorySamba Shiva, Ashrayahttp://hdl.handle.net/1893/356762024-01-26T12:16:07Z2023-09-30T00:00:00ZTitle: Modeling hippocampal theta-coupled gamma oscillations in learning and memory
Author(s): Samba Shiva, Ashraya
Abstract: Two of the most researched domains in the hippocampus are the oscillatory activity and
encoding and retrieval of patterns in the hippocampal CA1 and CA3 regions. They are, how-
ever, not studied together; and hence, the objective of our work is to study the cross-frequency
coupling of theta-coupled gamma oscillations in CA1 and CA3 regions of the hippocampus
while encoding and retrieving information. We have studied the cross-frequency coupling
of theta-coupled gamma oscillations both individually and in our newly-proposed integrated
model of CA1-CA3 to analyze the effects of Schaffer collaterals and CA1 back-projection
cells on CA1 and CA3 regions of the hippocampus. Due to lack of literary evidence, we have
also contributed our hypotheses about the effects of CA1 back-projection cells on CA1 and
CA3 cell-types. Moreover, we have developed a deterministic rule-based cellular automata
library to study cross-frequency coupling in single-neuron level and population neuronal net-
works at the same time. The discrete model is theta-oscillations-aware and hence encoding
and retrieving of patterns takes place during the half-cycles of theta oscillations.
We have extended the septo-hippocampal population firing rate model proposed by Den-
ham and Borisyuk (2000) to study (i) the influence of inhibitory interneurons, specifically
PV-containing basket cells (BCs) and bistratified cells (BSCs) on theta and theta-coupled
gamma oscillations in both CA1 and CA3 hippocampal networks; (ii) to study Schaffer col-
laterals from CA3 to CA1 and the influence of back-projection cells in CA1 on CA3; (iii) to
analyze and compare the phases of cross-frequency coupling of theta-coupled gamma oscil-
lations among the different cell types in CA1 and CA3 regions; (iv) to study the influence of
external inputs on CA1 and CA3. In our simulations, with constant external inputs, we identify the parameter regions that
generate theta oscillations and that BCs and BSCs in CA1 are in anti-phase, as seen experi-
mentally by Klausberger et. al (2008). Slow-gamma oscillations are generated due to the ac-
tivity of BSCs and BCs in CA1 and CA3, and they are propagated from CA3 to CA1 through
the Schaffer collaterals, as seen in Klausberger et. al (2008) where BSCs were observed to
synchronize PC activity during theta-coupled gamma oscillations in CA1. In CA3, increas-
ing excitation of CA3 pyramidal cells results in theta oscillations without the slow-gamma
coupling. Increasing excitatory input to CA1 pyramidal cells results in steady state and de-
creasing the excitatory input, results in reduced oscillatory activity in both CA1 and CA3 due
to Schaffer collaterals and the feedback projections from CA1 to CA3. This demonstrates
that changes in input excitation can move the networks from oscillatory to non-oscillatory states, comparable to the differences seen in animals between exploratory and resting state.
Further, Mizuseki et. al (2009) observed experimentally that CA1, CA3 and EC are out-
of-theta-phase with each other and that the phase observed in CA1 pyramidal cells are not a
result of a simple integration of phases from CA3, EC or the medial septum. We have thus,
simulated theta-frequency sine-wave inputs from CA3 and EC of relative phases in the model
and observed the same results in our CA1 individual and CA1-CA3 integrated model.
To study encoding and retrieval of patterns in an oscillating model, we took an engineer-
ing approach by developing a discrete modeling system using cellular automata (CA) derived
from the models of Pytte et. al (1991) and Claverol et. al (2002). The aim of this model is
to (i) replicate the oscillatory and phasic results obtained using the continuous modeling ap-
proach and (ii) extend the same model to study storage and recall of patterns in CA1 taking a
theta-oscillations-aware approach.
Encoding and retrieval happen at different half-cycles of theta where information pro-
cessing takes places in the sub-cycles of the slow-gamma oscillations in each half-cycle of
theta oscillation (Cutsuridis et. al, 2010, Hasselmo et. al, 1996). A set of rules is developed
to replicate this for the CA model of CA1. The encoding and retrieval half-cycles are identi-
fied using the basket cell activity, and hence synaptic learning is enabled during the encoding
half-cycle of theta, and is disabled during the recall half-cycle of theta oscillations. This is
also a biologically realistic enhancement for studying learning and recall in theta-coupled
gamma oscillations using a discrete cellular automata approach.2023-09-30T00:00:00ZOptimising WLANs Power Saving: Context-Aware Listen IntervalSaeed, Ahmedhttp://hdl.handle.net/1893/355362023-11-15T09:54:12Z2023-09-05T00:00:00ZTitle: Optimising WLANs Power Saving: Context-Aware Listen Interval
Author(s): Saeed, Ahmed
Abstract: Energy is a vital resource in wireless computing systems. Despite the increasing popularity of Wireless Local Area Networks (WLANs), one of the most important outstanding issues remains the power consumption caused by Wireless Network Interface Controller (WNIC). To save this energy and reduce the overall power consumption of wireless devices, a number of power saving approaches have been devised including Static Power Save Mode (SPSM), Adaptive PSM (APSM), and Smart Adaptive PSM (SAPSM). However, the existing literature has highlighted several issues and limitations in regards to their power consumption and performance degradation, warranting the need for further enhancements.
This thesis proposes a novel Context-Aware Listen Interval (CALI), in which the wireless network interface, with the aid of a Machine Learning (ML) classification model, sleeps and awakes based on the level of network activity of each application. We focused on the network activity of a single smartphone application while ignoring the network activity of applications running simultaneously.
We introduced a context-aware network traffic classification approach based on ML classifiers to classify the network traffic of wireless devices in WLANs. Smartphone applications’ network traffic reflecting a diverse array of network behaviour and interactions were used as contextual inputs for training ML classifiers of output traffic, constructing an ML classification model. A real-world dataset is constructed, based on nine smartphone applications’ network traffic, this is used firstly to evaluate the performance of five ML classifiers using cross-validation, followed by conducting extensive experimentation to assess the generalisation capacity of the selected classifiers on unseen testing data. The experimental results further validated the practical application of the selected ML classifiers and indicated that ML classifiers can be usefully employed for classifying the network traffic of smartphone applications based on different levels of behaviour and interaction.
Furthermore, to optimise the sleep and awake cycles of the WNIC in accordance with the smartphone applications’ network activity. Four CALI power saving modes were developed based on the classified output traffic. Hence, the ML classification model classifies the new unseen samples into one of the classes, and the WNIC will be adjusted to operate into one of CALI power saving modes. In addition, the performance of CALI’s power saving modes were evaluated by comparing the levels of energy consumption with existing benchmark power saving approaches using three varied sets of energy parameters. The experimental results show that CALI consumes up to 75% less power when compared to the currently deployed power saving mechanism on the latest generation of smartphones, and up to 14% less energy when compared to SAPSM power saving approach, which also employs an ML classifier.2023-09-05T00:00:00ZAudit Scotland: Improving audit quality with data scienceRichardson, Vikkihttp://hdl.handle.net/1893/355272023-11-08T14:36:34Z2022-12-01T00:00:00ZTitle: Audit Scotland: Improving audit quality with data science
Author(s): Richardson, Vikki
Abstract: Audit Scotland is appointed by the Auditor General for Scotland and the Accounts Commission to perform audit services for most of Scotland's public organisations. An auditor must determine if the accounts presented by an organisation represent a true and fair view of their financial position. A detailed, methodical exploration of the audit client's general ledger will assist the auditor in coming to this conclusion. Advances in the private audit sector in general ledger analysis have been made possible thanks to the streamlined nature of many small/medium enterprise (SME) accounts packages and the accounting frameworks that apply. Public audit financial management systems are more diverse in nature, as are the accounting frameworks that apply across the sectors, therefore no commercially available ledger analysis tools have been successfully adapted for use in public audit. Audit Scotland had introduced a rudimentary ledger analysis tool, in the form of a Microsoft Excel add-in, which was struggling to cope with the volume of data to be processed for effective analysis. It also failed to meet standards for reproducibility and documentation required by data analytics tools used in an audit.
The introduction of an ETL (extract, transform and load) data pipeline using data engineering principles during this project has increased Audit Scotland's capacity to ingest and prepare general ledger data from public organisations, ready for analysis. Additionally, a web application for general ledger analysis (Asc) has been created to enable Audit Scotland financial auditors transparent, efficient access to the general ledger data of their clients.
An important aspect of audit work is journal risk assessment. Each journal in the ledger should be assessed and classified as 'risky' or 'non-risky' in terms of causing a material misstatement in the accounts. With some public organisations creating upwards of a million journals per year this is an impossible manual task for an Audit Scotland audit team. Attempts at producing unsupervised machine learning classification models for this task by the author, thus far, have proved unsuccessful in accurately classifying journals thanks, in no small part, to an ineffective evaluation method caused by a lack of labelled data and a lack of resources needed to evaluate the model’s output. Within Asc, a journal risk assessment module has been developed which allows an auditor to manually classify all journals more efficiently through the lens of optional risk factors. This human expert classification has been captured to produce a data labeller for public audit ledger data, opening future possibilities to train and test supervised classification models, using labelled data, whilst providing a useful journal risk assessment tool for use in Audit Scotland now.
Using data visualisation tools and automated reporting within Asc, the general ledger can be examined to a significantly higher level of transparency. Auditors are reporting more confidence in their audit decisions, and the evidence they can provide to support those decisions, using Asc. Increased confidence, backed up by documented decisions lead to a higher quality audit which is the motivation for this work.2022-12-01T00:00:00ZIntegration of improved clustering and routing protocol in wireless sensor networks with crowd management systemsAlharbi, Mohammad Alihttp://hdl.handle.net/1893/354602023-10-11T12:49:58Z2022-09-01T00:00:00ZTitle: Integration of improved clustering and routing protocol in wireless sensor networks with crowd management systems
Author(s): Alharbi, Mohammad Ali
Abstract: Billions of devices are connected through wireless networks with utility in many areas of applications such as health, logistics, banking, smart homes, and smart cities. These devices can work in groups to achieve common objectives. However, these devices have limitations of residual battery power. One of the solutions is clustering in which nodes are grouped in which one node acts as Cluster-Head (CH) and other nodes as Cluster-Members (CMs). The responsibility of CH is to collect data from CMs and forward it to the Base-Station (BS). However, the transmission range of devices is limited and for large areas, multi-hop communication is required to collect data at BS. The solutions proposed in the literature either consider clustering or routing only. The literature review indicates a lack of unified clustering and routing protocol for the transmission of data to central BS.
In this research, fixed-area-based equal clustering and routing protocol is proposed where nodes inside a given area form a cluster. These nodes select one node as CH and other nodes act as CMs. CH is selected based on the accumulative weight of the residual battery, node degree, and node-centrality of nodes. In the CH selection process, the next-hop forward and backward nodes are identified for establishing the routing path.
Equal clustering has the drawback where the nodes close to the BS drain their batteries earlier due to the dual responsibility of data collection and forwarding. To solve this, an unequal-area-based clustering and routing solution is proposed where the cluster size near BS is small with fewer nodes and it increases as moves away from the center of the circular area. BS is deployed at the center of the area. The results of unequal clustering are better than that of equal clustering.
The case study of Hajj is considered an application of the proposed protocols. In the scenario of Tawaf, a large number of pilgrims circulate the Kaaba a fixed number of times. The number of clusters and nodes inside each cluster is computed using the pilgrim's data. Initially, an abnormal behavior detection model for a large number of pedestrians is developed. This model is applied to sensor data and extended to sensor-based solutions operating under the proposed unequal clustering and routing protocol. It is identified that the sensor-based solution provides better performance gain in terms of network lifetime, abnormal behavior detection time, and network throughput. The study further identified limitations in existing research and proposed future research areas.2022-09-01T00:00:00Z