Completed MS and PhD Theses
Comparison of Return Rate Efficiencies of Forecasting Methods in Stock Market Investment
Um_alkher Saaed Meina, February 2017
Prediction of prices in stock market is an important research topic to direct investments to items with high return rates. This thesis compares available time series prediction methods for predicting of stock market prices. The available methods that have been employed for time series forecasting are support vector regression, autoregressive moving average and k-nearest neighbours. They are applied on four years of stock market data obtained from London Stock Exchange to train each model and to test the performance of the proposed techniques to select the best forecasting method. The result of the tests show that support vector regression gives less forecasting error compared to other methods of forecasting.
Keywords: Stock Market Forecasting, Support Vector Regression, ARMA, k-Nearest Neighbours
Effect of Temporal Filters on Face Images
Rasheed Rebar Ihsan, February 2017
Face detection from low-resolution videos is a challenging research area. This thesis explores the effect of a temporal filtering method by Dr. Bodur on face images. The temporal mean and median filters calculate the intensity of pixels using the intensities of surrounding neighbour pixels, and temporal neighbour pixels in consecutive images. The effect of the proposed technique on the image is measured by the mean square error (MSE) and the peak signal noise ratio (PSNR) values using the pixels of the original high resolution image as reference values to measure the error and noise figures of the pixels of filtered low resolution images. Results demonstrate a significant effect of the proposed filters on the consecutive frames of face video record. In the tests, the median filter is found more effective compared to mean filter.
Keywords: Temporal Mean Filter, Temporal Median Filter, Image Resolution, Effectiveness of Image Filter.
Enhancement of Vehicle License Plate Images by Temporal Filtering
Diler Naseradeen Abdulqader, February 2017
Optical Character recognition is used widely as a tool in intelligent transportation systems for recognition of the car license plate from a still image or video. The accuracy of Optical Character Recognition partially depends on the quality of the input image. In this study, a set of simple and efficient methods are proposed to improve the quality of the car license plate image extracted from video clips to reduce the error rate for the license plate OCR even at low resolutions. Mean, median, and maximum filters are commonly used algorithms to filter noise and enhance an image. The proposed technique by Dr. Bodur extends them to time domain by including the pixels of the consequent images of the video clip in filtering algorithm. The OCR error rate is tested on fifty road and street video clips by decreasing the resolution of the images and filtering them with common and proposed filtering methods. The test results indicate that all proposed methods, improve the accuracy of OCR, and the highest reduction of error is obtained by the proposed temporal maximum filtering method
Keywords: License Plate Recognition, temporal image enhancement, Vehicle Plate OCR.
Temporal Ultrasound Image Enhancement for Kidney Diagnosis
Pawan Shivan Othman, February 2017
Ultrasound imaging enables the physician to view the tissues and organs in abdominal region of the body without hazards of ionization compared to the other radiation based internal organ inspection devices. It provides highly accurate imaging of a kidney on suspected acute renal diseases. This thesis proposes temporal filtering methods to enhance the ultrasound images from ultrasound kidney video for the purposes of identification and diagnosis of kidney diseases by processing consecutive images of the acquired kidney video extending the spatial mean, median and weighted mean image filters to temporal dimension after cropping and aligning the image frames manually in MATLAB by Image Processing Toolbox to suppress speckle noise, and improve information content for a diagnosis by a medical doctor. The assessment of the filtered images by 10 medical experts indicates that the proposed temporal mean, median, and weighted mean filters improve the images better than the common spatial mean, median and weighted mean filters. The evaluators ranked the temporal weighted mean filtered images as the least preferable, while they scored the temporal median filtered images as the best preferable ultrasound images for the purpose of renal diagnosis.
Keywords: Ultrasound Kidney Image, Image Enhancement by Video, Temporal Mean filter, Temporal Median Filter, Temporal Weighted Mean Filter
Development of Topological Mappings for Autonomous Agricultural Vehicles
Moein Mehrolhassani, Spring 2015-16
Automation system of agricultural crop plantation requires many subsystems such as low level tracking, path planning, obstacle detection, manoeuvres at the path terminations, etc. This study proposes semantic annotation for the information flow between the automation subsystems, filling the gap between the planning and implementation of crop production by developing two missing subunits: determination of obstacles that may threaten agricultural vehicles using the satellite images of target field, and determination of proper path for the agricultural vehicles to process rows of crops. For the attributes of obstacles, semantic annotation on the map of target field is preferred using Resource Description Framework/Extensible Mark-up Language (RDF/XML) in order to be exchangeable and reusable with other stages, systems, devices and applications. Developed Matlab code determines the target field by a GPS coordinate inside the field. An interactive initialization stage provides download of the satellite images from Google Maps API for determination of the field boundaries. The code for detection and positioning of the circular shaped obstacles are using Prewitt, Sobel, Roberts, and Canny edge detection, and Hough transformation algorithms. The developed method is tested on 51 target fields. It provides 45 % improvement in detection error rate compared to raw application of the algorithms.
Visualization of 3D Object on Planar Screen Using View Angle
Zuhir A. Badr, Spring 2014-15
This thesis developed and demonstrated a practical method to support 3D perception of stationary objects in a virtual space through the motion of a two dimensional projection image. The structure of a human eye is naturally equipped by some tools to perceive the depth from several hints such as the size of image compared to the its expected size, and the sharpness of the image at different focal lengths of the lens, the parallax difference in the images from the left and right eyes, and, if the image moves, by comparing the images at different view angles. In this thesis, the movement of the observer is detected by a software using the video camera frames, and the expected 2D projection of the virtual objects is transformed for the detected position of the observer to support a depth feeling of the observer. The developed program is coded in MATLAB, to determine the position of a red marker that is attached to the head of the observer, to compose the transformation matrix that converts 3D corner points of the virtual objects to expected perspective projection for the determined view-angle, and to draw the projection on the screen for the observation. The code is written in a flexible form to work with any PC with a web-cam, and graphical screen. The implemented system is tested successfully comparing the views of a set of virtual geometric objects on a platform with respect to the view of similar objects physically on a test platform.
Design, Development, and Prototyping of an Intelligent Volumetric Measurement System for Water Containers
Minoo Ekrani, Fall 2014-15
The aim of this thesis is to build an intelligent water volume measurement device that calculates the water volume for any shape of tank precisely by measuring the water level with an ultrasonic sensor and using a look-up-table to calculate the volume. It generates alarm signals at any desired volume. For any shape of tank, the look-up-table is entered to the processors memory using five buttons, and LCD display. The prototype gets power from mains through a USB socket even if it is not connected to a computer. The system is designed at a system-level using basic building blocks for Arduino from a Arduino Uno processor board, a keypad + LCD display shield with a LED, an ultrasonic HC-SR04 sensor unit, and an external buzzer. The prototype is tested successfully for a horizontally placed cylindrical container to have maximum 1.3%, average 0.7% error on distance and percent volume measurements.
Multi-objective Optimization of LARP Parameters using Weighted Sum DE Method
Behnam Seyedi, Summer 2013-14
This thesis employed the single-objective differential evolution (DE) algorithm to search the multi-objective solutions to obtain lateral controller (LARP) settings for an auto-steered tractor by combining two fitness functions, lateral peak and RMS errors, to a single objective using the weighted-sum-method. Compared to the multi-objective differential algorithm, weighted-sum DE algorithm covered a larger range of the Pareto-front. After modifying DE to set the search space adaptively, the modified method finds better non-dominated solutions than MODE by less number of fitness evaluations. Weighted Sum DE algorithm obtained better non-dominated solutions than MODE algorithm although weighted sum DE uses 20 000 fitness evaluation while MODE used 500 000 evaluations.
Comparison of the Edge Detection Methods to Detect, Identify and Locate the Obstacles for Agricultural Robotic Vehicles.
Anas Qasim Mahdi, Fall 2013-14
The obstacle detection in an agricultural field is an important step of the automation of the plantation. There are already developed autonomous agricultural vehicles that can track a path, and perform the specified processes on the plantation fields. These autonomous agricultural robotic machines need an upper level of control, which is mostly performed manually, for the design of the reference paths. Detection of the agricultural obstacles is necessary to accomplish these manual tasks in an automatic manner. In this study, statistical methods are employed to determine which of the five well-known edge-detection methods is best, for the high-level path planning in an agricultural automation of autonomous agricultural vehicles depending on field and image properties.
MS Thesis:Multi Objective Optimization of control parameters for Auto-steering System of Off-Road Vehicles.
Hamed Mahdizadeh, Spring 2012-13
In this thesis the search of optimum control parameters of an automated farming vehicle is planned to track a predetermined path on a loose soil surface. The problem is already formulated and a suboptimal solution is obtained in the PhD thesis “Design and Development of an Auto Steering System, Control for Off-Road Vehicles” by E. Kiani. The optimization requires to minimize both RMSE and Peak Lateral Error of tracking on the linear and circular paths, and it is suitable to a multi objective optimization approach. The solution of this problem might be important for the control of the Automated Farming Vehicles, since the reduction of the lateral error provides higher efficiency in automation of farming processes.
Prediction of Time Series Datasets using Fuzzy Functions.
Mahammad Abdulrazzaq Thanoon
In this thesis the past five year data of four international stock markets will be analyzed using the fuzzy time series analysis methods to compare the performance of statistical methods that provides the highest utility and the lowest risk for financial investment among the four major international stock markets: NewYork-US, Tokyo, London, and ShangHai, against the advanced fuzzy prediction methods. The fuzzy time series prediction is a recently emerging research field in artificial intelligence, and the results of the analysis can be used by Financial operators for finalizing their decision to shift values between these four major stock markets.
Model Based Multi objective Decision Making Methods for prediction of time series data
Ahmed Salih Ibrahim
In this thesis the estimation of the global stock market data is targeted using the last five year time series data of global stock market prices and volumes. The method of the prediction will be based on combining more than one common methods through fuzzy modeling methods to have higher prediction accuracy and higher profit/risk ratio. The thesis will combine available time series prediction methods such as Bollinger moving average, ARMA or ARIMA, Radial Basis Function Modeling, and Genetic algorithms applied on ARMA.
PhD Thesis: Design and Implementation of an Auto-Steering System Control for Off-Road Vehicles
Ehsan Kiani, September 2012
Abstract of Abstract In this thesis, the lateral error of a farm tractor at the
curvature transitions is minimized by introducing
a second look-ahead reference point (LARP) to the conventional lateral deviation controller. ... Present study develops a simple
automatic path tracking system to satisfy the typical requirements of an
unmanned agricultural tractor application based on properties of two look-ahead reference points (LARPs) on the desired path. The
main objective of the proposed control system is
to track a desired path within reasonable tolerances of a typical farming process including considerable slippage. ... the employed look-ahead reference points provide compensation for centrifugal forces and reduction of the peak lateral deviation due to curvature transition, using only simple arithmetic operations. ... Simulation results indicate enhancements in vehicle manoeuvrability and reduction of peak lateral displacement error at the curvature transitions one fifth of single LARP error. The proposed 2-LARP control strategy performs exactly same as the conventional lateral deviation controller on the linear and circular paths but it outperforms the conventional controller at the curvature transitions where the second LARP behaves independent to the first LARP.
Keywords: look-ahead reference point (LARP) control, path tracking, automatic steering agricultural vehicle, curvature transition.
Title: Prediction of Stock Market Values using Statistical Time Series Analysis Methods.
Jehan Kadhim Shareef, September 2013
In this thesis the past five year data of four international stock markets will be analyzed using the financial time series analysis to determine the best statistical method that provides the highest utility and the lowest risk for financial investment among the four major international stock markets: NewYork-US, Tokyo, London, and ShangHia. The time series prediction problem is an important paradigm in artificial intelligence, and the results of the analysis can be used by Financial operators for finalizing their decision to shift values between these four major stock markets.
MS Thesis: Symbolic Kinematics and Dynamics Modeling of a Four Wheel Mobile Inverted Pendulum
Asemeh Pousti, September 2009
Abstract: This thesis developed the basic kinematics and dynamics expressions for a Four Wheel Mobile Inverted Pendulum (FWMIP), which is made of a rod mounted on a four wheel non-holonomic mobile cart vehicle similar to ordinary automobiles. Our initial motivation was to develop a control strategy to keep the rod in the upward position meanwhile the vehicle is moving in a horizontal plane. The behavior of the system is explained in homogenous coordinates in terms of Robotics Kinematics rules. We obtained the symbolic equation of motion of FWMIP in terms of Robotics Dynamics rules that is suitable for any control algorithm. All the steps of finding the Kinematics and Dynamics model and control strategy of FWMIP were developed by using MATLAB toolbox. Closed loop feedback control combined with a PID controller algorithm is proposed for the 3-DOF system, and simulated successfully for the control of the planar motion of the system.
A Comparison of Fuzzy Functions with LSE and TS-Fuzzy Modeling Methods in Modeling Uncertain Datasets
Baharak Ahmaderaghi, September 2009
Abstract: This thesis compares the prediction performance of two Fuzzy modeling methods: Turksen’s Fuzzy Function with Least Squares Estimation (FF-LSE) and the Takagi-Sugeno’s (TS) Fuzzy Model. TS model is also known as a local linear modeling because of multidimensional linear consequents in its rule base. In Turksen’s model fuzzy sets are used as the multidimensional functions at the antecedent part of the fuzzy rules. According to the test results with five benchmark datasets FF-LSE gives in average 10 percent less error than TS modeling. FF-LSE needs less computational effort for both to construct the model and to infer.