Texas A&M University-Kingsville: AKM Digital Repository
Not a member yet
1689 research outputs found
Sort by
Bitcoin prediction and trading using machine learning
No full textThis thesis attempts to improve upon models that use Blockchain features to predict the
future value of Bitcoin. The features are compared to the price each time step and features
having a correlation above a set threshold are used as inputs to the model. Multiple models,
including linear, ridge, lasso, RANSAC, SVR, Theil-Sen, random forest, and gradient boosting
regression, are all run with this technique and then compared. This approach leads to improved
error compared to other models that do not change their variables over time. The predictions
from these models are also used to make trades and compared to conventional trading strategies.
Genetic algorithms were used to find the optimal parameters for the trading strategies. The
improved training method led to lower error on all models. However, the trading strategies
performed better than the prediction models when making trades
Effects of super absorbent polymer-soil interaction on plant growth, development, water use efficiency and soil leachate
No full textWater and nutrient availability are the most limiting factors in crop production. The rising cost of fertilizers and unavailability of water for irrigation are major concerns to farmers. Crop production in South Texas is particularly challenging because of the area’s semi-arid climate and silty clay loam soil type that requires constant irrigation to increase crop optimum growth and yield. Previous studies show that superabsorbent polymers (SAP) have been used as soil amendments. They absorb water and nutrients and release them slowly in the soil so the nutrients are available for root absorption for increased crop growth, development and yield. A type of SAP (ZEBA®) was used in this study at concentrations used in the Netherlands at Wageningen Plant Research, Vredepeel to study ZEBA’s interaction with south Texas soil, and to determine its optimum concentration for soils in the area. Results indicate that ZEBA mixed with south Texas soils at 1:200 and 1:500 showed no statistical difference in all growth parameters studied. Also, no difference exists between the two concentrations in leachate volume, water retention capacity, and nitrate concentration of leachate. However, Capsicum annum grown in these soil concentrations of ZEBA had significantly higher chlorophyll content (P=0.0103) relative to ambient south Texas soil with or without fertilizer. Soil media mixed with ZEBA used significantly less water (p=0.0158), less leachate (p=0.0007) and nitrate in leachate (p<0.0001) compared to soils of south Texas
Non-linear finite element buckling analysis of steel column piles
No full textThis research paper examines nonlinear buckling of structural steel I-shaped columns used as piles in coastal regions. The piles are 360in long I-shaped cross-sections made of structural steel. The selected cross-sections are: W14x82 with compact flange and web, W14x90 with a non-compact flange and compact web, and W14x120 with non-compact flange and web. Each cross-section will be modeled four times with varying degrees of section loss due to corrosion (0%, 25%, 50% and 75% loss of the cross-section). Cross-section losses will occur along a splash zone.
The research goal is to develop analytical models, carry out a linear analysis to get the eigenvalue of the first buckling mode shape of the model, then conducting a non-linear post buckling analysis using the Modified Riks method. This paper intends to consider both material and geometric nonlinearity in carrying out the post-buckling analysis. ABAQUS/Standard was used to carry out the linear and non-linear computational testing on shell extrusion type models. The deformation, maximum principal stress, critical buckling, and failure loads are recorded. The buckling mode shapes are compared to theoretical results, and the critical buckling loads are validated with the Euler equation. The results from the post-buckling analysis are also compared with nominal compressive strength, Pn, as calculated with equations in the AISC manual. Results showed good agreement with the AISC failure load calculations.
This thesis could be used by engineers and scientists to better estimate the remaining strength capacities and performance of steel columns that have lost a section due to natural phenomena. The results could also help predict real life responses of structural steel members under different boundary conditions and loading
Optimized parallel algorithms for computing PageRank
No full textPageRank is the measure of importance of a node within a set of nodes. It was originally developed for ranking pages in the web to provide relevant and latest search results for a user’s query. However, it is very useful and is still being used in many other domains such as social network analysis, spam detection and bibliometrics. Due to the continuous growth of real-world networks at a remarkable speed, massive graphs present a significant challenge for search engines. These graphs can contain millions or even billions of nodes and edges, which demands the need for efficient parallel algorithms. In this thesis, we design an efficient and optimized parallel algorithm for computing PageRank.
There exists several good parallel algorithms for computing PageRank, however, they still suffer from communication overhead and load imbalance. Our algorithm provides improved and optimized communication strategy, better load balance and higher speed up. We developed and incorporated an efficient communication mechanism which reduces inter processor communication significantly. This has led to communication efficiency of 2-4-fold over the existing algorithm for the 7 graphs we have experimented with. We have also incorporated a parallel load balancing algorithm which provides faster partitioning mechanism and improved load balancing. Our experiments on several real-world datasets show that the proposed algorithm provides superior performance over the existing algorithms
Microalgae biomass production: a case of municipal wastewater remediation and microalgae harvesting
No full textCurrently wastewaters are explored for algal biomass production not only to produce biomass for biofuels production and co-products, but also to significantly remove a myriad of contaminants present in wastewater. The microalgae biomass generation for biofuels from this source is yet to be commercialized. Microalgae growth and biochemical compositional parameters, such as biomass concentration and productivity, lipid, pigments, carbohydrate, and protein productivities, are usually evaluated for economic feasibility of biomass generated by manipulating the growth conditions such as altering the growth nutrient compositions, temperature, salinity and mixing intensity among others. Another concern of commercializing microalgae biomass generation for biofuels is high energy demanding dewatering process of obtaining biomass paste from very dilute, negatively charged, stable suspension of microalgae cells in culture broth.
The first objective of this dissertation was to evaluate the growth, contaminants removal, and biochemical component (lipid, carbohydrate and protein) accumulation potential of Chlorella vulgaris, Scenedesmus obliquus and Oocystis minuta cells in wastewater supplemented with different concentrations of sulfate, nitrate and phosphate. The experimental results in this case indicated maximum biomass productivities of 33, 19 and 98 mg dw/L/d for C. vulgaris, S. obliquus, and O. minuta, respectively. More than 90% phosphate removal was highest in the culture with O. minuta and C. vulgaris was able to remove about 93% nitrate, and sulfate removal (36%) was highest in the S. obliquus culture. The microalgae strain S. obliquus demonstrated the highest sulfate removal of 36%, followed by C. vulgaris (35%) and O. minuta with 27%. The biochemical composition of the microalgae cells was 22-65% carbohydrate, 19-38% protein and 8-17% lipid. This was an indication that carbohydrate and protein were favorably accumulated as compared to lipids, under the growth conditions investigated for each of the microalgae strains.
In the second objective, the potential of Dunaliella salina, Nannochloropsis oculata, Oocystis minuta, and Neochloris conjuncta to accumulate biomass, pigments, lipid, carbohydrate, and protein under different mechanical agitation of the culture systems were evaluated. The laboratory results showed that maximum biomass concentration (2,423 mg/L) and productivity (544 mg/L/d) were obtained in D. salina at a stirring speed of 1,400 rpm and an aeration rate of 0.15 vvm. Also, total lipid content of 23% was maximum for N. oculata at a shaking speed of 200 rpm. However, 49% protein and 45% carbohydrate content were recorded as maximum for D. salina and O. minuta at 125 and 200 rpm shaking speeds, respectively. Under the investigated mechanical agitation of the cultures, more protein and carbohydrate were obtained as compared to lipid for all microalgae species, and these were identified as better candidates for bioethanol production and protein source for animal feed, rather than biodiesel production.
Moreso, in the third objective, the flocculating potential of agricultural waste derived bio-based alkali from plantain peel ash was assessed. The flocculation results revealed that microalgae harvesting efficiencies were more than 90% for both Nannochloropsis oculata and Dunaliella salina at 1, 2 and 3% v/v flocculant (bio-based alkali) dose, while the maximum biomass concentration factors were 20 and 14 at 1% v/v flocculant dose for N. oculata and D. salina, respectively. The characterization results of bio-based alkali employed to induce microalgae flocculation showed that the concentration of K+ (61,630 ppm) was highest, followed by Ca2+ (1,569 ppm) and Na+ (582 ppm). This confirmed the alkaline nature of the bio-based used as flocculant. Therefore, the produced plantain peel-derived alkali can be used to effectively and efficiently harvest microalgae biomass from very dilute microalgae culture media.
Moreover, in the fourth objective, the D-optimal design of the response surface methodology (RSM) was employed to assess modeling, prediction and optimization of microalgae flocculation process by RSM, artificial neural network (ANN) and genetic algorithm (GA). The bio-based alkali was produced from cocoa pod husk ash and used to induce flocculation of microalgae. In this case, the results indicated the potential of bio-based alkali to successfully flocculate the microalgae with more than 90 % flocculation efficiencies at the optimum condition predicted by RSM and RSM-GA. Both RSM and ANN gave high accurate results of the flocculation process; this was indicated by statistical indices evaluated (R values closer to 1), but ANN demonstrated a higher generalization potential compared to RSM. The results of elemental analysis of the cocoa pod husk bio-based alkali revealed that the concentration of K+ (51,489 ppm) was highest, followed by Ca2+ (1,450 ppm) and Na+ (210 ppm). This undeniably showed the alkaline nature of the bio-based alkali obtained from cocoa pod husk ash that was used to induce the microalgae flocculation Therefore, this also confirmed that produced cocoa pod derived alkali can be used efficiently in harvesting microalgae from dilute microalga culture broth
Students', teachers', and coordinators' perceptions about students' use of code-switching in their English as foreign language classroom
No full textLearning English as a foreign language (EFL) had high relevance for many people around the world due to globalization, technology, personal, and professional goals. People considered English as the most important foreign language to be learned, because of the tourism industry, and the industrial development making English necessary to work in the northeast of Mexico, (Coupland, 2013; Maier, Siemund, & Davydova, 2012) but mostly because of Mexico´s border with the United States of America.
This research based cognitive psychology, sociolinguistics, and bilingualism as the theoretical framework to give an account of students´, teachers´, and coordinators´ perceptions about students´ use of code-switching in their EFL classroom, and to achieve English as a foreign language with proficiency. Therefore, the research questions were: (a) what are students´ perceptions about their use of code-switching in their EFL classroom? (b) what are teachers´ perceptions about students´ use of code-switching in their EFL classroom? and (c) what are coordinators´ perceptions about students´ use of code-switching in their EFL classroom? The participants were 193 students, 12 teachers, and 5 coordinators from two language centers that belonged to a northeast university in Mexico during the spring semester of 2019 and who consented to volunteer in the research process.
A quantitative cross-sectional survey design explored the participants´ code-switching perceptions at a certain point in the semester. Data gathering comprised of a self-administered web-based survey for students, teachers, and coordinators via Survey Monkey ®. Data collection was analyzed using the statistical program IBM SPSS ® version 21; also, tables with frequencies and percentages reflected the participants´ code-switching perceptions.
The quantitative data exposed the participants´ perceptions about code-switching in their EFL classrooms reflecting the following findings: (a) the participants supported code-switching practices in the EFL classroom, (b) there was a higher tendency of student-student code-switching practices rather than student-teacher code-switching practices, (c) most of the participants had a positive or neutral attitude about code-switching practices in the EFL classroom, and (d) participants´ perceptions reflected that students used code-switching as a learning tool to acquire, understand, learn, and communicate in English.
Keywords: English as a foreign language, code-switching, and perception
Android application development: a brief analysis of Android architecture, evaluation of security loopholes and development of preventive mechanisms
No full textNowadays, mobile phones are used as multipurpose devices that fit in the palm of the hand having the capability to take photos or sending emails. However, with the advent of mobile technologies, attackers are developing new techniques to bypass security mechanisms and breach into the security systems of mobile phones. In this report, a brief introduction to mobile viruses along with security principles to sustain the confidentiality, availability, and integrity of the data has been discussed. Previous research that proposed novel techniques of mobile security have been discussed. Moreover, layered and cross-platform approaches are discussed to understand Android architecture. Current statistics of attacks on mobile devices and points of future concern have been analyzed. A detailed analysis of malware and currently existing security loopholes in the Android operating system is presented. Nowadays, malicious applications can be developed by any moderate programmer. In this work, as a proof of concept, a malicious application is developed that can retrieve confidential data of users without their consent. Additionally, a preventive mechanism to encrypt credentials authorized by users has also been developed to prevent a data breach and provide users the flexibility to control their data settings and privacy
VHDL implementation of reduced complexity Wallace multiplier and performance analysis comparison
No full textOne of the most important core components for arithmetic operations in a digital system is a multiplier. The overall performance of a multiplier is dependent on adder performance. Wallace multiplier is one of first approaches to improve the performance of the traditional multiplier operation. After that, many modified approaches were proposed to improve the performance of Wallace multiplier and the other types of multipliers. Various types of adders are used to design the enhanced Wallace tree multiplier. One of the very recent introduced full adder is called energy efficient CMOS full adder (EECFA). The research area for improving overall multiplier performances is vast and the three most important matrices to determine the performance of a multiplier are area, speed and power dissipation. In this thesis, one of the most recently proposed, the reduced complexity Wallace multiplier using energy efficient CMOS full adder was implemented using the VHSIC Hardware Description Language and then a comparative performance analysis with the conventional Wallace Multiplier and Dadda Multiplier is presented in terms of area, speed and power dissipatio
Analysis of the change of wind farm layouts in Texas
No full textArrays of wind turbines spread over a vast area of land makes a wind farm, and wind energy is becoming the most popular renewable energy source. Wind farms have layouts, and their layouts vary in different locations, however there is no comprehensive understanding of the changes of wind farms’ layouts. In this study, Geographic Information System was utilized to evaluate 11,559 wind turbines installed in Texas from 2010 to 2019. Geospatial and temporal data analyses were performed with data gathered from different databases to obtain the outlines of wind farms and to understand the various patterns that affect them. The findings of the geospatial research were investigated further using a numerical analysis. With the use of Geographic Information Systems-based tools, different changes of the wind farms were compared and assessed including the maximum count of wind turbines in a wind farm by year, the number of total wind farms installed every year, the total area of a wind farm, wind turbine nameplate capacity, wind turbine hub height, and wind turbine rotor diameter. To recognize different wind speed categories at different locations, wind speed data was included in the analysis. The results showed that the number of new wind turbines and wind farms built every year do not have an obvious trend, but the hub height and rotor diameter show an increasing trend every year
Co-ordinates based graphical module for effective virtual pavement condition assessment
No full textThe involvement of technology in all the fields of engineering from the past decade has been increasing immensely. With the advent of computer data and interpretation models, civil engineering practices has broadened its spectrum in addressing contingencies and improving over human errors that occur in old practices. This thesis is an attempt to improve the conventional practices used by TxDOT in using google earth to survey and record distress data which correspondingly help in the management system and validate the pavement condition through images. The concentration of the research is towards the practice used in TxDOT’s maintenance and construction division in pavement management. In this research attempt a graphical user interface (GUI) is created using MATLAB which extracts the co-ordinate data from google earth and record the distress details. The information from the interface is used to create data and validate the pavement condition of the recorded year based on previous research concentrating in quick distress scanning and determining pavement condition. Comparing the pavement condition map for the current year with survey details of the recorded year, the effect of the results of maintenance activities, areas to be concentrated and issues with preventive measures used are discussed and verified