Engineering Commons^™

Forecasting Pedestrian Trajectory Using Deep Learning, Arsal Syed

UNLV Theses, Dissertations, Professional Papers, and Capstones

In this dissertation we develop different methods for forecasting pedestrian trajectories. Complete understanding of pedestrian motion is essential for autonomous agents and social robots to make realistic and safe decisions. Current trajectory prediction methods rely on incorporating historic motion, scene features and social interaction to model pedestrian behaviors. Our focus is to accurately understand scene semantics to better forecast trajectories. In order to do so, we leverage semantic segmentation to encode static scene features such as walkable paths, entry/exits, static obstacles etc. We further evaluate the effectiveness of using semantic maps on different datasets and compare its performance with already …

Mitigating Insider Threat Risks In Cyber-Physical Manufacturing Systems, Jinwoo Song

Dissertations - ALL

Cyber-Physical Manufacturing System (CPMS)—a next generation manufacturing system—seamlessly integrates digital and physical domains via the internet or computer networks. It will enable drastic improvements in production flexibility, capacity, and cost-efficiency. However, enlarged connectivity and accessibility from the integration can yield unintended security concerns. The major concern arises from cyber-physical attacks, which can cause damages to the physical domain while attacks originate in the digital domain. Especially, such attacks can be performed by insiders easily but in a more critical manner: Insider Threats.

Insiders can be defined as anyone who is or has been affiliated with a system. Insiders have knowledge …

Adaptive Mobile Eeg Noise Cancellation Using 2d Convolutional Autoencoders For Bci Authentication, Tyree Lewis

USF Tampa Graduate Theses and Dissertations

Electroencephalography (EEG) signals can be used for many purposes and has the potential to be adapted to various systems. When EEG is recorded from users, these studies are performed primarily in an indoor environment, while the user is stationary. This is due to the levels of noise that are experienced when recording EEG data, to minimize errors in the data. This thesis aims to adapt tasks that are performed indoors to an external environment by removing both noise and artefacts in EEG, using a 2D Convolutional Autoencoder (CAE). The data is recorded from subjects is passed into the 2D CAE …

Studying Complex Aquifer Systems From Large-Scale Stratigraphy Development To Local Aquifer Storage And Recovery, Hamid Vahdat Aboueshagh

LSU Doctoral Dissertations

Hydrostratigraphy model is an essential component of building valid groundwater models. Many challenges are associated with constructing hydrostratigraphy models which include geological complexities such as faults, domes, and angular unconformities. Developing a method with an emphasis on capturing big data to thoroughly inform large-scale models is one of the challenges addressed in the first part of this study. The method is predicated upon discretization of the study domain into tiles based on the geological dip direction and faults. The application of the method in the state of Louisiana with the utilization of more than 114000 well logs demonstrates promising results …

Data-Driven Studies On Social Networks: Privacy And Simulation, Yasanka Sameera Horawalavithana

USF Tampa Graduate Theses and Dissertations

Social media datasets are fundamental to understanding a variety of phenomena, such as epidemics, adoption of behavior, crowd management, and political uprisings. At the same time, many such datasets capturing computer-mediated social interactions are recorded nowadays by individual researchers or by organizations. However, while the need for real social graphs and the supply of such datasets are well established, the flow of data from data owners to researchers is significantly hampered by privacy risks: even when humans’ identities are removed, or data is anonymized to some extent, studies have proven repeatedly that re-identifying anonymized user identities (i.e., de-anonymization) is doable …

Stock Market Manipulation Detection Using Continuous Wavelet Transform & Machine Learning Classification, Sarah Youssef

Theses and Dissertations

Stock market manipulation detection is important for both investors and regulators. Being able to detect stock manipulation and preventing it gives investors the confidence in the market fairness and integrity. It also helps maintaining liquidity of the stocks and market efficiency. Implementing data mining algorithms in manipulation detection is a relatively recent technique but in the past few years there has been an increasing interest in it's applications in this domain. The benefit of monitoring manipulative trade behavior is that it can be implemented on live feed of stock data, which saves a lot of time in detecting stock price …

Design Of An Automatic Fish Processing Line Using Machine Learning, Sangam Mainali

Master's Theses

This thesis presents an approach to automatic fish processing in the fishing industry using artificial intelligence. Using transfer learning approach (VGG16 model), convolutional neural network (CNN), and image processing, this thesis demonstrates the series of steps from intake of fishes, chopping of head and tail to collecting the body parts of fishes separately thus making it ready for post-processing or direct packaging as per the industrial and consumer needs. To present the idea of the classification of fishes, we have considered five different fishes and one shrimp, altogether six classes. After the classification, the fishes are sorted out according to …

Observation Of The Evolution Of Hide And Seek Ai, Anthony J. Catelani

Computer Science and Software Engineering

The purpose of this project is to observe the evolution of two artificial agents, a ‘Seeker’ and a ‘Hider’, as they play a simplified version of the game Hide and Seek. These agents will improve through machine learning, and will only be given an understanding of the rules of the game and the ability to navigate through the grid-like space where the game shall be played; they will not be taught or given any strategies, and will be made to learn from a clean slate. Of particular interest is observing the particular playstyle of hider and seeker intelligences as new …

Artificial Intelligence In Real-Time Diagnostics And Prognostics Of Composite Materials And Its Uncertainties – A Review, Muthu Ram Prabhu Elenchezhian, Vamsee Vadlamudi, Rassel Raihan, Kenneth Reifsnider, Erick Reifsnider

UTARI Researcher Publications-Archive

In the era of the 4th industrial revolution of big data, Artificial Intelligence (AI) is widely used in each and every field of composite materials which includes design and analysis, material storage, manufacturing, non-destructive testing (NDT), Structural Health Monitoring (SHM) and Prognostics of its Remaining Useful Life (RUL), Material State (MS) and damage modes. While these AI models are rapidly developed and integrated into the Industrial Internet of Things (IIoT) to keep track of the health of a composite material from its birth to death, these integrations remain uncertain for prognostics without the certainty of its previous material state. This …

Applying Deep Learning On Financial Sentiment Analysis, Cuiyuan Wang

Dissertations, Theses, and Capstone Projects

Portfolio Investment has always been appealing to investors and researchers. In the past, people tend to use historical trading information of the securities to predict the return or manage the portfolio. Nowadays, the literature has been proved that the market sentiment could predict asset prices. Specifically, it has been shown that the stock market movement is related to financial news and social media events. Thus, it becomes necessary to extract the sentiment of the financial news. We explicitly introduce the application of dictionary methods, traditional machine learning models and deep learning models on text classification. The experiment results show that …

Implementing A Data Acquisition System For The Training Of Cloud Coverage Neural Networks, Weston C. Montgomery

Master's Theses

Cal Poly is home to a solar farm designed to nominally generate 4.5 MW of electricity. The Gold Tree Solar Farm (GTSF) is currently the largest photovoltaic array in the California State University (CSU) system, and it was claimed to be able to produce approximately 11 GWh per year. These types of projections come from power generation models which have been developed to predict power production of these large solar fields. However, when it comes to near-term forecasting of power generation with variable sources such as wind and solar, there is definitely room for improvement.

The two primary factors that …

Soarnet, Deep Learning Thermal Detection For Free Flight, Jake T. Tallman

Master's Theses

Thermals are regions of rising hot air formed on the ground through the warming of the surface by the sun. Thermals are commonly used by birds and glider pilots to extend flight duration, increase cross-country distance, and conserve energy. This kind of powerless flight using natural sources of lift is called soaring. Once a thermal is encountered, the pilot flies in circles to keep within the thermal, so gaining altitude before flying off to the next thermal and towards the destination. A single thermal can net a pilot thousands of feet of elevation gain, however estimating thermal locations is not …

Bibliometric Review Of Predictive Maintenance Using Vibration Analysis, Aashna Midha Ms., Ishita Maheshwari Ms., Kaushik Ojha Mr., Kritika Gupta Ms., Shripad V. Deshpande Mr.

Library Philosophy and Practice (e-journal)

Every day the world is depending more and more on machines in almost every aspect of life. With the increasing use of machines, there also needs to be an evolution in the maintenance of these machines. Predictive maintenance is a process used to monitor the equipment and machinery during its operation to detect any damages and/or deteriorations and enable the required maintenance plan in advance, resulting in reduced operational costs and full utilization of tools and parts. The fundamental goal of this bibliometric review paper is a comprehension of the extent and sources of the literature available for predictive maintenance …

Bibliometric Study On Analysing Impact Of Newly Launched Products Over Existing Ones Through Ai, Dipak Sharma, Vageesh Devrath, Abhinav Rajput, Gouranga Jyoti Kataky, Priyanka Tupe-Waghmare, Ismail Akbani

Library Philosophy and Practice (e-journal)

Different analysis models like Conditional Mean Analysis, Trend Analysis, Correlation Analysis helps us to analyse the delicate equilibrium between businesses that gets impacted when a new product is launched in a cluster. This paper shows a statistical report of research done on the businesses in a cluster based on ongoing trends and current customer needs . There is surplus data present on various platforms related to every product following the ongoing trends in the form of customer reviews.The research mainly speculates mainly how the businesses get impacted with change in consumer needs, wants and demands. With the help of datasets …

On Stability And Similarity Of Network Embeddings, Apurva Shriniwas Mulay

Theses - ALL

Machine Learning on graphs has become an active research area due to the prevailing graph-structured data in the real world. Many real-world applications can be modeled with graphs. Modern application domains include web-scale social networks [26], recommender systems, knowledge graphs, and biological or protein networks. However, there are various challenges. First, the graphs generated from such applications are often large. Moreover, in some scenarios, the complete graph is not available, e.g., for privacy reasons. Thus, it becomes impractical to perform network analysis or compute various graph measures. Hence, graph sampling becomes an important task.Sampling is often the first step to …

On Stability And Similarity Of Network Embeddings, Apurva Shriniwas Mulay

Theses - ALL

Machine Learning on graphs has become an active research area due to the prevailing graph-structured data in the real world. Many real-world applications can be modeled with graphs. Modern application domains include web-scale social networks [26], recommender systems, knowledge graphs, and biological or protein networks. However, there are various challenges. First, the graphs generated from such applications are often large. Moreover, in some scenarios, the complete graph is not available, e.g., for privacy reasons. Thus, it becomes impractical to perform network analysis or compute various graph measures. Hence, graph sampling becomes an important task.Sampling is often the first step to …

Real-Time Monitoring Of Fdm 3d Printer For Fault Detection Using Machine Learning: A Bibliometric Study, Vaibhav Kisan Kadam, Satish Kumar, Arunkumar Bongale

Library Philosophy and Practice (e-journal)

Additive Manufacturing has wide application range including healthcare, Fashion, Manufacturing, Prototypes, Tooling etc. AM techniques are subjected to various defects that may be printing defects or anomalies in machine. There is gap between current AM techniques and smart manufacturing since current AM lacks in build sensors necessary for process monitoring and fault detection. Both of these issues can be solved by incorporating real-time monitoring into AM. So the study is carried out to identify recent work done in AM to improve current system. For this bibliometric study Scopus database is used, study is kept limited to year 2010-2021 and English …

Bibliometric Survey On Flood Prediction Using Machine Learning, Seema Patil Prof., Daksh Khurana Mr., Kartik Rao Mr, Priyanshu Meena Mr, Shivendra Singh Mr

Library Philosophy and Practice (e-journal)

Floods are one of the most devastating natural hazards, and modelling them is extremely difficult. Flood prediction model advancement study led to factors such as loss of human and animal life, property damage, and risk mitigation. The focus of this bibliometric survey is to recognise the few studies which have upheld on the factors affecting the floods. The analysis is done based on 254 documents such as articles, conference papers, article reviews and some reviews and notes. India contributes to the maximum number of documents followed by China and the United States of America. This bibliometric survey is conducted using …

Prediction Of Stocks And Stock Price Using Artificial Intelligence : A Bibliometric Study Using Scopus Database, Priyanka Tupe-Waghmare, Priyanka Tupe-Waghmare

Library Philosophy and Practice (e-journal)

Prediction of stocks and the prices of the stock is one of the most crucial points of discussion amongst the researchers and analysts in the financial domain to date. Every stakeholder and most importantly the investor desires to earn higher profit for his investment in the market and try to use several different strategies to invest their money. There are numerous methods to predict and analyse the movement of the stock prices. They are broadly divided into – statistical and artificial intelligence-based methods. Artificial intelligence is used to predict the futuristic prices of stocks and use wide range of algorithms …

Machine Learning Applied To Colloidal Properties Of Perfluorocarbon Nanoemulsions For Imaging In Ards/Ali, Marco Hosfeld

Electronic Theses and Dissertations

Acute Respiratory distress Syndrome (ARDS) and Acute Lung Injury (ALI) are inflammatory lung pathologies consisting of non-hydrostatic pulmonary edema leading to hypoxia and impaired gas exchange in the lungs. ARDS/ALI is both difficult to study and treat as it is not in itself a specific pathology but rather a syndrome consisting of many pathologies that vary case by case. It is, however, consistently characterized by an explosive acute inflammatory response in the lung parenchyma leading to hypoxia. Although time has seen to an increase in the understanding of ARDS/ALI, the mortality rate remains in the range of 30-50%. For these …

Decentralized Aggregation Design And Study Of Federated Learning, Venkata Naga Surya Sameeraja Malladi

Master of Science in Software Engineering Theses

The advent of machine learning techniques has given rise to modern devices with built-in models for decision making and providing rich content to users. This typically involves processing huge volumes of data in central servers and sending updated models to end-user devices. There are two main concerns on this server architecture, one is the privacy of data that is being transferred to a central server and the other is volumes of data sent over the network for the model update. Federated Learning helps solve these problems by training models on local data within the device and aggregating the model with …

Field Demonstration Of Gpr And Uav Technologies For Evaluation Of Two Us 75/77 Bridges, Sepehr Pashoutani, Jinying Zhu, Chungwook Sim, Ji-Yong Lee

Nebraska Department of Transportation: Research Reports

Two Nebraska bridges with asphalt overlay were selected for nondestructive testing and evaluation (NDT/NDE). Three NDT techniques were conducted on these two bridges, including Ground Penetrating Radar (GPR), Half-Cell Potential (HCP) and Unmanned Aerial Vehicle (UAV) imaging. NDT data were collected during three construction stages of the bridges: (1) before repair on existing asphalt overlay; (2) on bare concrete after asphalt removal; (3) and after repairing delaminated concrete.

A machine learning technique, autoencoder, was used to build quantitative relationships between different NDT datasets. On bare concrete, the GPR amplitude and HCP voltage show a strong linear relationship. Then a threshold …

Optimal Analytical Methods For High Accuracy Cardiac Disease Classification And Treatment Based On Ecg Data, Jianwei Zheng

Computational and Data Sciences (PhD) Dissertations

This work constitutes six projects. In the first project, a newly inaugurated research database for 12-lead electrocardiogram signals was created under the auspices of Chapman University and Shaoxing People's Hospital (Shaoxing Hospital Zhejiang University School of Medicine). This database aims to enable the scientific community in conducting new studies on arrhythmia and other cardiovascular conditions. In the second project, we created a new 12-lead ECG database under the auspices of Chapman University and Ningbo First Hospital of Zhejiang University that aims to provide high quality data enabling detection of the distinctions between idiopathic ventricular arrhythmia from right ventricular outflow tract …

A Study Of Deep Reinforcement Learning In Autonomous Racing Using Deepracer Car, Mukesh Ghimire

Honors Theses

Reinforcement learning is thought to be a promising branch of machine learning that has the potential to help us develop an Artificial General Intelligence (AGI) machine. Among the machine learning algorithms, primarily, supervised, semi supervised, unsupervised and reinforcement learning, reinforcement learning is different in a sense that it explores the environment without prior knowledge, and determines the optimal action. This study attempts to understand the concept behind reinforcement learning, the mathematics behind it and see it in action by deploying the trained model in Amazon's DeepRacer car. DeepRacer, a 1/18^th scaled autonomous car, is the agent which is trained …

Synthesizer Parameter Approximation By Deep Learning, Daniel Faronbi, Alisa Gilmore

Theses/Capstones/Creative Projects

Synthesizers have been an essential tool for composers of any style of music including computer generated sound. They allow for an expansion in timbral variety to the orchestration of a piece of music or sound scape. Sound designers are trained to be able to recreate a timbre in their head using a synthesizer. This works well for simple sounds but becomes more difficult as the number of parameters required to produce a specific timbre increase. The goal of this research project is to formulate a method for synthesizers to approximate a timbre given an input audio sample using deep learning. …

Application Of Machine Learning In Flood Depth Prediction, Armando Esquivel

Open Access Theses & Dissertations

Machine learning technologies have helped provide answers for problems with a high degree of complexity. Machine learning has been utilized by various disciplines within the Civil Engineering profession and has proven to be efficient in solving complex problems. Although machine learning is being used in the Civil Engineering profession, a formal framework on developing and integrating machine learning has not been developed for flood depth prediction. The proposed word uses machine learning to predict the depth of flood at Houston, TX, due to a 100-year 24-hour storm. The proposed work can be used to collect, store and analyze data to …

Statistical Machine Learning For Breast Cancer Detection With Terahertz Imaging, Tanny Andrea Chavez Esparza

Graduate Theses and Dissertations

Breast conserving surgery (BCS) is a common breast cancer treatment option, in which the cancerous tissue is excised while leaving most of the healthy breast tissue intact. The lack of in-situ margin evaluation unfortunately results in a re-excision rate of 20-30% for this type of procedure. This study aims to design statistical and machine learning segmentation algorithms for the detection of breast cancer in BCS by using terahertz (THz) imaging. Given the material characterization properties of the non-ionizing radiation in the THz range, we intend to employ the responses from the THz system to identify healthy and cancerous breast tissue …

Analog Spiking Neural Network Implementing Spike Timing-Dependent Plasticity On 65 Nm Cmos, Luke Vincent

Graduate Theses and Dissertations

Machine learning is a rapidly accelerating tool and technology used for countless applications in the modern world. There are many digital algorithms to deploy a machine learning program, but the most advanced and well-known algorithm is the artificial neural network (ANN). While ANNs demonstrate impressive reinforcement learning behaviors, they require large power consumption to operate. Therefore, an analog spiking neural network (SNN) implementing spike timing-dependent plasticity is proposed, developed, and tested to demonstrate equivalent learning abilities with fractional power consumption compared to its digital adversary.

Characterizing Students’ Engineering Design Strategies Using Energy3d, Jasmine Singh, Viranga Perera, Alejandra Magana, Brittany Newell

Discovery Undergraduate Interdisciplinary Research Internship

The goals of this study are to characterize design actions that students performed when solving a design challenge, and to create a machine learning model to help future students make better engineering design choices. We analyze data from an introductory engineering course where students used Energy3D, an open source computer-aided design software, to design a zero-energy home (i.e. a home that consumes no net energy over a period of a year). Student design actions within the software were recorded into text files. Using a sample of over 300 students, we first identify patterns in the data to assess how students …

Five Degree-Of-Freedom Property Interpolation Of Arbitrary Grain Boundaries Via Voronoi Fundamental Zone Octonion Framework, Sterling Gregory Baird

Theses and Dissertations

In this work we introduce the Voronoi fundamental zone octonion (VFZO) interpolation framework for grain boundary (GB) structure-property models and surrogates. The VFZO framework offers an advantage over other five degree-of-freedom (5DOF) based property interpolation methods because it is constructed as a point set in a Riemannian manifold. This means that directly computed Euclidean distances approximate the original octonion distance with significantly reduced computation runtime (∼7 CPU minutes vs. 153 CPU days for a 50000×50000 pairwise-distance matrix). This increased efficiency facilitates lower interpolation error through the use of significantly more input data. We demonstrate grain boundary energy (GBE) interpolation results …