Engineering Commons^™

Deep Learning (Partly) Demystified, Vladik Kreinovich, Olga Kosheleva

Departmental Technical Reports (CS)

Successes of deep learning are partly due to appropriate selection of activation function, pooling functions, etc. Most of these choices have been made based on empirical comparison and heuristic ideas. In this paper, we show that many of these choices -- and the surprising success of deep learning in the first place -- can be explained by reasonably simple and natural mathematics.

Towards A Theoretical Explanation Of How Pavement Condition Index Deteriorates Over Time, Edgar Daniel Rodriguez Velasquez, Carlos M. Chang Albitres, Vladik Kreinovich

Departmental Technical Reports (CS)

To predict how the Pavement Condition Index will change over time, practitioners use a complex empirical formula derived in the 1980s. In this paper, we provide a possible theoretical explanation for this formula, an explanation based on general ideas of invariance. In general, the existence of a theoretical explanation makes a formula more reliable; thus, we hope that our explanation will make predictions of road quality more reliable.

Nonlinear Mechanical Properties Of Road Pavements: Geometric Symmetries Explain The Empirical Difference Between Roads Built On Clay Vs. Granular Soils, Afshin Gholamy, Vladik Kreinovich

Departmental Technical Reports (CS)

It is empirically known that roads built on clay soils have different nonlinear mechanical properties than roads built on granular soils (such as gravel or sand). In this paper, we show that this difficult-to-explain empirical fact can be naturally explained if we analyze the corresponding geometric symmetries.

Computer-Aided Classification Of Impulse Oscillometric Measures Of Respiratory Small Airways Function In Children, Nancy Selene Avila

Open Access Theses & Dissertations

Computer-aided classification of respiratory small airways dysfunction is not an easy task. There is a need to develop more robust classifiers, specifically for children as the classification studies performed to date have the following limitations: 1) they include features derived from tests that are not suitable for children and 2) they cannot distinguish between mild and severe small airway dysfunction.

This Dissertation describes the classification algorithms with high discriminative capacity to distinguish different levels of respiratory small airways function in children (Asthma, Small Airways Impairment, Possible Small Airways Impairment, and Normal lung function). This ability came from innovative feature selection, …

Cmas/Volcanic Ash Infiltration Performance Of Yttria Rich-Zirconia Thermal Barrier Coatings Produced By Electron Beam Physical Vapor Deposition, Juan Jose Gomez Chavez

Open Access Theses & Dissertations

For many years the use of thermal barrier coatings (TBC) in gas turbine engines has allowed a significant increase in engine operating temperatures which are above the Ni-based super alloys' melting point. The base material for TBCs is the state of the art 7 wt % yttria stabilized zirconia (YSZ) which provides the thermal protection. Current industry demands for increased engine efficiency has pushed TBCs to the limits where significant technological constrains have been identified. One of the main ones represents the high temperature corrosion of the TBC material due to the ingestion of silicate based airborne particles commonly referred …

Towards Analytical Techniques For Systems Engineering Applications, Griselda Valdepeñas Acosta

Open Access Theses & Dissertations

One of the main objectives of systems engineering is to design, maintain, and analyze systems that help the users. To design an appropriate system for an application domain, we need to know: what are the users' desires and preferences (so that we know in what direction we should aim to change this domain), what is the current state and what is the dynamics of this application domain, and how to use all this information to select the best alternatives for the system design and maintenance. Designing a system includes selecting numerical values for many of the parameters describing the corresponding …

Strategies To Improve The Electrochemical Of Performance Transition Metal Compounds As Anode Materials For Li-Ion Batteries, Noemi Dominguez

Open Access Theses & Dissertations

Nowadays, our society depends on fossil fuels as the main energy source. However, depletion, price fluctuations, and other concerns, such as global warming, have led to an exhaustive research for renewable energy resources. Here is where electrochemical energy systems play a critical role since they can provide power when renewable energy is not available. Among these electrochemical energy storage systems Li-ion batteries (LIBs) have shown higher volumetric, gravimetric, and power density capacities over other battery systems.

However, commercial LIBs are still relying on graphite base materials for the anode. However, graphite possesses a low specific capacity. Recently, transition metal compounds …

Computational Studies On Perovskite-Metallofullerene Interface And Magnetic Properties Of Mn-Based Mixed Olivines, Bethuel Omutiti Khamala

Open Access Theses & Dissertations

Methyl ammonium lead halide (MAPbX3) perovskite based solar cells have recently emerged as promising class of materials for photovoltaic applications with efficiencies reaching over 22%. Designing interfaces with strong binding is vital to developing efficient, high-performing solar cells. Fullerene-based materials are widely employed as efficient electron acceptors and can serve as electron transporting layer in perovskite based solar cells. We have studied interfaces of methyl ammonium lead iodide MAPbI3 with Sc3N@C80 fullerene and Sc3N@C80PCBM fullerene derivate within the density functional formalism. Different surface terminations and orientations of the methyl ammonium are examined for binding of the fullerene layer. Our calculated …

Formulation And Implementation Of Iterative Method For Generating Spatially-Variant Lattices, Manuel Fernando Martinez

Open Access Theses & Dissertations

The use of a matrix-free, memory-efficient approach to generate large-scale spatially variant lattices (SVL) was explored. A matrix-free iterative SVL generation algorithm was formulated and then implemented with a tremendous memory reduction observed. The algorithm consists of solving first-order central finite-differences along the entirety of the problem space point-by-point to obtain the grating phase function Φ(𝑠⃗) to which all desired spatially variant lattice properties are applied to. The algorithm was studied to identify key areas of data and task parallelism to exploit in heterogeneous computing systems consisting of clusters of central processing units (CPU) and graphics processing units (GPU) combinations. …

Synthesis, Characterization And Antibacterial Properties Of Silver(I) N-(4-Pyridyl) Benzamide Complexes With Varying Conterions, Mauro J. Ruiz

Open Access Theses & Dissertations

Organic semiconductor materials composed of - stacking aromatic compounds have been under intense investigation for their potential uses in flexible electronics and other advanced technologies. Herein we report a new family of seven - stacking compounds of Silver(I) N-(4-pyridyl) Benzamide with varying counterions, namely [Ag(NPBA)2]X, where NPBA is N-(4-pyridyl) Benzamine, X = NO3- (1), ClO4- (2), CF3SO3- (3), PF6- (4), BF4- (5), CH3PhSO3- (6), and PhSO3- (7), which form extended - stacking networks in one-dimensional (1D), 2D and 3D directions in the crystalline solid-state via the phenyl moiety, with average inter-ring distances of 3.823 Å. Interestingly, the counterions that contain …

Dedicated Hardware For Machine/Deep Learning: Domain Specific Architectures, Angel Izael Solis

Open Access Theses & Dissertations

Artificial intelligence has come a very long way from being a mere spectacle on the silver screen in the 1920s [Hml18]. As artificial intelligence continues to evolve, and we begin to develop more sophisticated Artificial Neural Networks, the need for specialized and more efficient machines (less computational strain while maintaining the same performance results) becomes increasingly evident. Though these “new” techniques, such as Multilayer Perceptron’s, Convolutional Neural Networks and Recurrent Neural Networks, may seem as if they are on the cutting edge of technology, many of these ideas are over 60 years old! However, many of these earlier models, at …

Magnetic Nanoparticles For Hyperthermia For Cancer Treatment, Bianca Paola Meneses Brassea

Open Access Theses & Dissertations

Fe3O4 and NixCu4-x magnetic nanoparticles were synthesized using supercritical conditions of liquids and wet chemistry, respectively. Characterization methods (VSM, SEM, TEM, and magnetic hyperthermia) yielded results that prove feasibility for magnetic hyperthermia for cancer treatment.

Energy Efficiency And Fatigue Failure On Gears And Bearings For Processing Rebar Steel, Jesus Gabriel Garcia

Open Access Theses & Dissertations

Steel industry is a major emitter of air pollutants, a leading industrial consumer of water, a minor source of contaminated liquids, a massive producer of solid waste, and a significant source of CO2. The objective on this thesis work is to minimize the greenhouse gas emissions when processing steel. Gears and bearings were evaluated based on their performance when exposed to cyclic stresses. The importance for having a proper set up on rolling mills to increase efficiency for processing rebar steel has been detailed on paper. In contrast, mechanical vibration has been measured on gears and bearings to determine efficiency …

Potential Of Nanoscale Elements To Control Fusarium Wilt Disease In Tomato (Solanum Lycopersicum), Enhance Macronutrient Use Efficiency, And Increase Its Yield, Ishaq Olarewaju Adisa

Open Access Theses & Dissertations

Nanotechnology has a great potential in ensuring food production, security and safety globally. Over the past decade, research on the use of nanomaterials to supply nutrient elements and protect plants from pest and diseases has significantly increased. Tomato (Solanum lycopersicum) is one of the most consumed vegetables in the world and United State is one of its largest producers globally generating billions of dollars annually in revenue.. Tomato plants are affected worldwide by Fusarium wilt caused by Fusarium oxysporum f. sp. Lycopersici. There is growing concern about excessive use of conventional pesticides in controlling Fusarium and other diseases in tomato …