Engineering Commons^™

Can Machine Learning Beat Physics At Modeling Car Crashes?, Gavin Byrne

Dissertations

This study aimed to look at a traditional method used for measuring the severity and principle direction of force of a car crash and see if it could be improved on using machine learning models. The data used was publicly available from the NHTSA database and included descriptions of the vehicle, test and sensors as well as the accelerometer data over the period of the crashes. The models built were SVM classifiers and multinomial regression models. Although the SVM and Regression models were built successfully and gave higher levels of accuracy than the momentum models in terms of the severity, …

Supervised Learning Models To Predict Stock Direction Within Different Sectors In A Bull And Bear Market, Tiffany Razy

Dissertations

Forecasting stock market price movement is a well researched and an alluring topic within the machine learning and financial realm. Supervised machine learning algorithms such as Random Forest (RF) and Support Vector Machines (SVM) have been used independently to gain insight on the market. With such volatility in the market the scope of this study will utilized the RF and SVM in a very volatility market to determine if these models will perform at a high level or outperform each other in both markets. This relative study is performed on 16 stocks in 4 different sectors over the bear market …

Identifying Expert Investors On Financial Microblog Via Artificial Neural Networks, Pierluca Del Buono

Dissertations

In the recent years, thanks to social media platform, a plethora of information has been available to financial investors, that were traditionally dependent from financial institutions advisors. Strategies are now shared among web users, performances of stocks are commented in web communities and hints and suggestions are travelling on the internet with a fast pace, in a way that was unthinkable few years before. Several attempts have been made in the recent past, to predict Market movements and trends from activity of Financial Social Networks participants, and to evaluate if contributions from individuals with high level of expertise distinguish themselves …

Bs News January/February

Building Services Engineering

No abstract provided.

Experienced Mental Workload, Perception Of Usability, Their Interaction And Impact On Task Performance, Luca Longo

Articles

No abstract provided.

Investigating The Application Of Deep Convolutional Neural Networks In Semi-Supervised Video Object Segmentation, Jayadeep Sasikumar

Dissertations

This thesis investigates the different approaches to video object segmentation and the current state-of-the-art in the discipline, focusing on the different deep learning techniques used to solve the problem. The primary contribution of the thesis is the investigation of usefulness of Exponential Linear Units as activation functions for deep convolutional neural architectures trained to perform object semi-supervised segmentation in videos. Mask R-CNN was chosen as the base convolutional neural architecture, with the view of extending the image segmentation algorithm to videos. Two models were created, one with Rectified Linear Units and the other with Exponential Linear Units as the respective …

Is It Worth It? Budget-Related Evaluation Metrics For Model Selection, Filip Klubicka, Giancarlo Salton, John D. Kelleher

Conference papers

Projects that set out to create a linguistic resource often do so by using a machine learning model that pre-annotates or filters the content that goes through to a human annotator, before going into the final version of the resource. However, available budgets are often limited, and the amount of data that is available exceeds the amount of annotation that can be done. Thus, in order to optimize the benefit from the invested human work, we argue that the decision on which predictive model one should employ depends not only on generalized evaluation metrics, such as accuracy and F-score, but …

Whispering-Gallery Mode Microsphere Resonators For Applications In Environmental Sensing, Arun Mallik

Doctoral

Humidity, temperature and volatile organic compounds (VOCs), particularly ammonia, are key environmental conditions that have a major impact on human comfort, well-being and productivity, as well as on agriculture, food processing and storage, electronic manufacturing and many other industries. This results in the urgent need for the development of sensing technologies allowing rapid detection and accurate measurement of these environmental parameters. Over the past decades many electrical as well as optical sensors have been proposed and demonstrated for environmental applications. However, challenge always exists for these sensors in terms of sensitivity, selectivity, detection limit, speed of response and robustness, where …

Human Performance Modelling For Adaptive Automation, Maria Chiara Leva, M. Wilkins, F. Coster

Conference papers

The relentless march of technology is increasingly opening new possibilities for the application of automation and new horizons for human machine interaction. However there is insufficient scientific evidence on human factors for modern socio-technical systems supporting the guidelines currently used to design Human Machine Interfaces (HMI) (ISA 2014). This dearth of knowledge presents a particular risk in safety critical industries. The continuing 60–90% of accidents currently that are rooted in Human Factors (HF) and the rapid developments in the Internet of Things (IoT) and its novel automation archetypes means that the requirements for new interfaces are becoming more demanding, and …

The Efficacy Of Financial Futures As A Hedging Tool In Electricity Markets, Jim Hanly, Lucia Morales, Damien Cassells

Articles

Abstract This paper estimates and applies a risk management strategy for electricity spot exposures using futures hedging. We apply our approach to three of the most actively traded European electricity markets, Nordpool, APXUK and Phelix. We compare both optimal hedging strategies and the hedging effectiveness of these markets for two hedging horizons, weekly and monthly using both Variance and Value at Risk (VaR). Our key finding is that electricity futures can effectively manage risk only for specific time periods when using hedging strategies that have been very successful in financial and other commodity markets. More generally they are ineffective as …

Examining A Hate Speech Corpus For Hate Speech Detection And Popularity Prediction, Filip Klubicka, Raquel Fernandez

Conference papers

As research on hate speech becomes more and more relevant every day, most of it is still focused on hate speech detection. By attempting to replicate a hate speech detection experiment performed on an existing Twitter corpus annotated for hate speech, we highlight some issues that arise from doing research in the field of hate speech, which is essentially still in its infancy. We take a critical look at the training corpus in order to understand its biases, while also using it to venture beyond hate speech detection and investigate whether it can be used to shed light on other …

Beef Cattle Instance Segmentation Using Mask R-Convolutional Neural Network, Mohammad Danish

Dissertations

Maintaining the cattle farm along with the wellbeing of every heifer has been the major concern in dairy farm. A robust system is required which can tackle the problem of continuous monitoring of cows. the computer vision techniques provide a new way to understand the challenges related to the identification and welfare of the cows. This paper presents a state-of-art instance segmentation mask RCNN algorithm to train and build a model on a very challenging cow dataset that is captured during the winter season. The dataset poses many challenges such as overlapping of cows, partial occlusion, similarity between cows and …

Video Quality Prediction Under Time-Varying Loads, Obinna Izima, Ruairí De Fréin, Mark Davis

Conference papers

We are on the cusp of an era where we can responsively and adaptively predict future network performance from network device statistics in the Cloud. To make this happen, regression-based models have been applied to learn mappings between the kernel metrics of a machine in a service cluster and service quality metrics on a client machine. The path ahead requires the ability to adaptively parametrize learning algorithms for arbitrary problems and to increase computation speed. We consider methods to adaptively parametrize regularization penalties, coupled with methods for compensating for the effects of the time-varying loads present in the system, namely …

High Sensitivity Optical Fiber Interferometric Sensors, Dejun Liu

Doctoral

Optical fiber interferometers have been widely employed and investigated for monitoring the changes in both physical and chemical parameters, with the advantages of compact size, light weight, immunity to electromagnetic interference, high sensitivity, capability to work in harsh environments and remote operation capabilities. Among the different kinds of fiber sensors based on interferometry, singlemode-multimode-singlemode (SMS) structures has attracted considerable interest due to their inherent advantages of high sensitivity, ease of fabrication and interconnection to other fiber systems and low cost. However, the challenge is that the sensitivity of the traditional SMS based fiber structure is not sufficient in some cases, …

From The Workshop To The Classroom - Examining Apprentice Perceptions, Anne Marie Mccarrick, Audrey Mccann, Odran Reid

Conference Papers

From the Workplace to the Classroom – Examining Apprentice Perceptions

The Irish government has reiterated its commitment to apprenticeships and how these form a central plank of the Further Education and Training Strategy. One of the main goals of the Action Plan for Education is to ‘create a greater diversity of learning opportunities beyond school’. Indeed, an increase of 19% in apprenticeship registrations took place between 2015 and 2016, and it is planned that this will grow by more than 30% by the year 2020.

This paper aims to identify both positive and negative perceptions of apprentices’ experiences as they …

Measurements Of Milli-Newton Surface Tension Forces With Tilted Fiber Bragg Gratings, Changyu Shen, Chuan Zhong, Dejun Liu, Xiokang Lian, Jianyao Zheng, Jingjing Wang, Yuliya Semenova, Gerald Farrell, Jacques Albert, John F. Donegan

Articles

Small lateral forces (lower than 0.1 N) cannot normally be measured with conventional single-mode fiber-based sensors because of the high value of their Young modulus (>70 GPa). Here we demonstrate the measurement of lateral forces in the range from 0.2 to 1.4 × 10−3 N with a tilted fiber Bragg grating (TFBG) in conventional single-mode fiber pushed against the surface tension (ST) of a bead of water. The measured transmission changes of individual cladding mode resonances of the TFBG corresponding to these force values are of the order of 29 dB. Separate measurements of the contact angle between the …

Optical Fibre Sensors For Monitoring Phase Transitions In Phase Changing Materials, Rahul Kumar, Wei Han, Dejun Liu, W. P. Ng, Binns, Richard Binns, Richard, Krishna K. Busawon, Yong Qing Fu, Zabih Ghassemlooy, Chris P. Underwood, Khamid Kh Mahkamov, Jinhui Yuan, Chongxiu Yu, Huazhong Shu, Xing Ao Li, Tuan Guo, Gerald Farrell, Yuliya Semenova, Qiang Wu

Articles

A platinum coated singlemode-multimode (SM) structure is investigated in this paper as an optical fibre sensor (OFS) to monitor the phase transition of a phase change material (PCM). Paraffin wax has been used as an example to demonstrate the sensor's performance and operation. Most materials have the same temperature but different thermal energy levels during the phase change process, therefore, sole dependency on temperature measurement may lead to an incorrect estimation of the stored energy in PCM. The output spectrum of the reflected light from the OFS is very sensitive to the bend introduced by the PCM where both liquid …

Formation And Development Of Nanometer-Sized Cybotactic Clusters In Bent-Core Nematic Liquid Crystalline Compounds, Yuri Panarin, Sithara P. Sreenilayam, Jagdish K. Vij, Anne Lehmann, Carsten Tschierske

Articles

Two homologue achiral bent-core liquid crystals (LCs), BCN66 and BCN84, in their nematic phases are studied by dielectric spectroscopy in the frequency range 10 Hz–10 MHz. In each of these compounds, two relaxation processes are identified and assigned to (i) collective dynamics of molecules in nanometer-sized cybotactic clusters and (ii) individual molecular relaxations, in the ascending order of frequency of the probe field. The temperature and the bias electric field dependence of the dielectric strength and relaxation frequency for these processes are shown to give rise to sharpness in cluster boundaries, increased size and volume fraction in the LC nematic …

The Role Of The Dendritic Growth Models Dimensionality In Predicting The Columnar To Equiaxed Transition (Cet), M. Seredyński, Marek Rebow, J. Banaszek

Articles

The dendrite tip kinetics model accuracy relies on the reliability of the stability constant used, which is usually experimentally determined for 3D situations and applied to 2D models. The paper reports authors` attempts to cure the situation by deriving 2D dendritic tip scaling parameter for aluminium-based alloy: Al-4wt%Cu. The obtained parameter is then incorporated into the KGT dendritic growth model in order to compare it with the original 3D KGT counterpart and to derive two-dimensional and three-dimensional versions of the modified Hunt’s analytical model for the columnar-to-equiaxed transition (CET). The conclusions drawn from the above analysis are further confirmed through …

An Lcl-Filter Design With Optimum Total Inductance And Capacitance, Sampath Jayalath, Mohamed Moin Hanif

Articles

LCL-filter is among the best performing filters for grid-connected voltage source inverters. Designing of the filter parameters (grid-side and inverter-side inductors and capacitor), takes an iterative approach due to the coherence between the parameters and design requirements such as IEEE-519 Std for harmonic current limitations, reactive power compensation limit, and maximum allowable voltage drop across the filter to limit the switching losses. Most of the proposed LCL-filter optimization strategies emphasize on reducing the total inductance and losses of the filter while meeting the design requirements. There is less emphasis given on the capacitor selection and optimizing its value. Therefore, this …

A Coated Spherical Micro-Resonator For Measurement Of Water Vapor Concentration At Ppm Levels In Very Low Humidity Environments, Arun Mallik, Gerald Farrell, Dejun Liu, Vishnu Kavungal, Qiang Wu, Yuliya Semenova

Conference Papers

This paper presents a novel approach to measurements of low relative humidity (RH) values based on the whispering gallery modes (WGMs) effect in a small silica microsphere coated with a thin layer of Agarose hydrogel. The light from a narrow-linewidth tunable laser with a few GHz tuning range is launched into an adiabatic tapered fiber, which excites WGMs in the Agarose layer via evanescent coupling. Adsorption and desorption of water vapor by the Agarose layer in response to changes in ambient humidity lead to changes in the layer’s refractive index, which in turn leads to a spectral shift of the …

Task Complexity, And Operators’ Capabilities As Predictor Of Human Error: Modeling Framework And An Example Of Application, Maria Chiara Leva, Alberto Caimo, Rebecca Duane, Micaela Demichela, Lorenzo Comberti

Conference Papers

This paper presents the initial framework adopted to assess human error in assembly tasks at a large manufacturing company in Ireland. The model to characterize and predict human error presented in this paper is linked conceptually to the model introduced by Rasch (1980), where the probability of a specified outcome is modelled as a logistic function of the difference between the person capacity and item difficulty. The model needs to be modified to take into account an outcome that is not dichotomous and feed into the interaction between two macro factors: (a) Task complexity: that sum-marises all factors contributing to …

Quantitative Fine-Grained Human Evaluation Of Machine Translation Systems: A Case Study On English To Croatian, Filip Klubicka, Antonio Toral, Victor Manuel Sanchez-Cartagena

Articles

This paper presents a quantitative fine-grained manual evaluation approach to comparing the performance of different machine translation (MT) systems. We build upon the well-established Multidimensional Quality Metrics (MQM) error taxonomy and implement a novel method that assesses whether the differences in performance for MQM error types between different MT systems are statistically significant. We conduct a case study for English-to- Croatian, a language direction that involves translating into a morphologically rich language, for which we compare three MT systems belonging to different paradigms: pure phrase-based, factored phrase-based and neural. First, we design an MQM-compliant error taxonomy tailored to the relevant …

Design Of Event-Triggered Fault-Tolerant Control For Stochastic Systems With Time-Delays, Yi Gao, Yunji Li, Li Peng, Junyu Liu

Articles

This paper proposes two novel, event-triggered fault-tolerant control strategies for a class of stochastic systems with state delays. The plant is disturbed by a Gaussian process, actuator faults, and unknown disturbances. First, a special case about fault signals that are coupled to the unknown disturbances is discussed, and then a fault-tolerant strategy is designed based on an event condition on system states. Subsequently, a send-on-delta transmission framework is established to deal with the problem of fault-tolerant control strategy against fault signals separated from the external disturbances. Two criteria are provided to design feedback controllers in order to guarantee that the …

Energy Efficient Hybrid Routing Protocol Based On The Artificial Fish Swarm Algorithm And Ant Colony Optimisation For Wsns, Xinlu Li, Brian Keegan, Fredrick Mtenzi

Articles

Wireless Sensor Networks (WSNs) are a particular type of distributed self-managed network with limited energy supply and communication ability. The most significant challenge of a routing protocol is the energy consumption and the extension of the network lifetime. Many energy-efficient routing algorithms were inspired by the development of Ant Colony Optimisation (ACO). However, due to the inborn defects, ACO-based routing algorithms have a slow convergence behaviour and are prone to premature, stagnation phenomenon, which hinders further route discovery, especially in a large-scale network. This paper proposes a hybrid routing algorithm by combining the Artificial Fish Swarm Algorithm (AFSA) and ACO …

Using Moocs To Promote Digital Accessibility And Universal Design, The Moocap Experience, John Gilligan, W. Chen, J. Darzentas

Articles

The recently completed Massive Open Online Course for Accessibility Partnership project (MOOCAP), had the twin aims of establishing a strategic partnership around the promotion of Universal Design and Accessibility for ICT professionals and of developing a suite of Open Educational resources (OERs) in this domain. MOOCAP's eight university partners from Germany, Norway, Greece, Ireland, the UK and Austria have a significant history in developing and providing courses in the domains of Universal Design and Accessibility, as well as leading research and advocacy roles within Europe. The MOOCAP project consisted of two phases: the development of an introductory MOOC on Digital …

Examining The Effects Of A Virtual Character On Learning And Engagement In Serious Games, Vihanga Gamage, Cathy Ennis

Articles

Virtual characters have been employed for many purposes including interacting with players of serious games, with a purpose to increase engagement. These characters are often embodied conversational agents playing diverse roles, such as demonstrators, guides, teachers or interviewers. Recently, much research has been conducted into properties that affect the realism and plausibility of virtual characters, but it is less clear whether the inclusion of interactive agents in serious applications can enhance a user’s engagement with the application, or indeed increase efficacy. In a first step towards answering these questions, we conducted a study where a Virtual Learning Environment was used …

An Investigation Of Argumentation Theory For The Prediction Of Survival In Elderly Using Biomarkers, Lucas Middeldorf Rizzo, Ljiljana Majnaric, Pierpaolo Dondio, Luca Longo

Conference papers

Research on the discovery, classification and validation of biological markers, or biomarkers, have grown extensively in the last decades. Newfound and correctly validated biomarkers have great potential as prognostic and diagnostic indicators, but present a complex relationship with pertinent endpoints such as survival or other diseases manifestations. This research proposes the use of computational argumentation theory as a starting point for the resolution of this problem for cases in which a large amount of data is unavailable. A knowledge-base containing 51 different biomarkers and their association with mortality risks in elderly was provided by a clinician. It was applied for …

On The Reliability, Validity And Sensitivity Of Three Mental Workload Assessment Techniques For The Evaluation Of Instructional Designs: A Case Study In A Third-Level Course, Luca Longo

Conference papers

Cognitive Load Theory (CLT) has been conceived for instructional designers eager to create instructional resources that are presented in a way that encourages the activities of the learners and optimise their performance, thus their learning. Although it has been researched for many years, it has been criticised because of its theoretical clarity and its methodological approach. In particular, one fundamental and open problem is the measurement of its cognitive load types and the measurement of the overall cognitive load of learners during learning tasks. This paper is aimed at investigating the reliability, validity and sensitivity of existing mental workload assessment …

On The Application Of Pspice For Localised Cloud Security, Paul Tobin

Doctoral

The work reported in this thesis commenced with a review of methods for creating random binary sequences for encoding data locally by the client before storing in the Cloud. The first method reviewed investigated evolutionary computing software which generated noise-producing functions from natural noise, a highly-speculative novel idea since noise is stochastic. Nevertheless, a function was created which generated noise to seed chaos oscillators which produced random binary sequences and this research led to a circuit-based one-time pad key chaos encoder for encrypting data. Circuit-based delay chaos oscillators, initialised with sampled electronic noise, were simulated in a linear circuit simulator …