Department of Computer Engineering

Permanent URI for this collectionhttp://197.255.125.131:4000/handle/123456789/4854

Browse

Search Results

Now showing 1 - 10 of 10
  • Thumbnail Image
    Item
    Efficient Modelling of a PCB Transmission Line for High Speed Digital Systems
    (2018) Acakpovi, A.; Sowah, R.; Asabere, N.Y.; et al.
    This paper proposes a model of PCB traces for high-speed digital systems. The adopted approach involves predetermined geometry using direct discretization of transmission lines. Initially, the proposed methodology involves computing the line propagation delay by employing its geometry with associated empirical equations. The initial procedure paves the way to design a Lattice diagram which depicts multiple reflections that the signal underwent due to impedance mismatches between transmission lines and loads. Subsequent computations of electrical model parameters were further done. Simulation results using Multisim software illustrated a favorable performance with a time delay of 1.42 ns and an equivalent electrical model of 10 lumped LC cells. The time delay between the input and output signal obtained from the simulation was approximately 15.152 ns corresponding to the time it took for a transmitted signal to reach a steady state which further signifies the good performance of our proposed method.
  • Thumbnail Image
    Item
    Self-Reporting System For Incidents Detection In Automated Teller Machine (Atm) Using Machine Learning Techniques
    (University Of Ghana, 2021-09) Payne, I.N.
    Automated Teller Machines (ATMs) have increased over the past decade due to their advantages in the banking sector. ATMs provide convenience to customers, optimizes banking operations, and minimizes transaction cost. However, undesirable security incidents such as tempering, skimming, physical attacks, robbery, and transaction reversal fraud may occur on ATM systems and negatively affect the user experience and banking institutions. ATM incidents occur either by system defect or through a deliberate act of physical attack by an intruder. In most security incidents, financial losses are imminent, and the customers' confidence in banking reduces. Developing a Self-Reporting System for ATM Incident Detection (SRSAID) is needed to avert the threats posed by security incidents on ATM systems. This research uses a machine-learning approach to solve this problem. Regional Convolutional Neural Network (R-CNN) and Support Vector Machine (SVM) algorithms are used to develop a detection model that detects occurrences of security incidents on an ATM system. Datasets used in the machine learning model development were obtained from NCR Ghana and the online repository. Experimental results showed that two CNN architecture models, ALEXNET and ssdlite_mobilenet_V2, obtained an accuracy score of 80% and 96%, respectively. SVM classifiers were developed using the linear, polynomial, and radial basis kernels, getting accuracy scores of 70.6%, 72.56%, and 81.21%, respectively. The initial results necessitated hyperparameter optimization to improve the performance of the classifiers. This resulted in improved accuracy scores of 76%, 77%, and 86% for linear, polynomial, and radial basis kernels, respectively, for the SVM models. The machine learning model was later deployed on a Raspberry Pi system which connected to a web application that provided a graphical user interface for user interactivity and viewing of reports.
  • Thumbnail Image
    Item
    Hybrid Cluster-Based Sampling Technique for Class Imbalance Problems
    (University of Ghana, 2019-07) Kuditchar, B.
    Class imbalance problem is prevalent in many real-world domains and as such has become an area of increasing interest for many researchers. In binary classification problems, imbalance learning refers to learning from a dataset with a high degree of skewness to the negative class. This phenomenon causes traditional classification algorithms to perform woefully when predicting positive classes with new examples. Data resampling is among the most commonly used techniques used to deal with this problem. It involves the manipulation of the training data before applying standard classification techniques. This study presents a new hybrid sampling technique that has the capability of improving the overall performance of a wide range of traditional machine learning algorithms. The proposed method uses an undersampling technique based on CUST to under-sample majority instances and an oversampling technique derived from SNOCC to oversample minority instances. The method is implemented in Python version 3.5 on windows. The performance was evaluated using classification algorithms from scikit-learn machine learning library, namely: KNN, SVM, Decision Tree, Random Forest, Neural Network, AdaBoost, Naïve Bayes, and Quadratic Discriminant Analysis. Eleven datasets with various degrees of imbalance were used. The performance of each classifier when the proposed technique is used is compared with the performance when no sampling is performed. In addition to that, the performance of eight (8) other sampling techniques is compared with that of the proposed method. These techniques include ROS, RUS, SMOTE, ADASYN, CUST, SBC, CLUS, and OSS. The experimental results showed that HCBST performed better with most of the classifiers in terms of AUC, G-Mean, and MCC. The overall average performance also showed that HCBT performed better in most of the datasets, having the highest average scores of 0.73, 0.67 and 0.35 in AUC, G-Mean and MCC respectively across all the classifiers used for this study. Extensive testing of machine learning algorithms and their performance metrics yielded promising results. A Graphical User Interface (GUI) to enable interactivity for machine learning with class imbalanced data task operations was incorporated to allow flexibility in the choice of algorithms for certain datasets for higher accuracy.
  • Thumbnail Image
    Item
    Adaptive Hybrid Collaborative Filtering Recommendation System (AHCF)
    (University of Ghana, 2019-07) Agboyi, R.
    Recommendation systems play a vital role in boosting the organization’s profit, especially for e-commerce platforms such as Amazon. These systems focus on targeting specific products to users and predicting user preferences and interests. However, recommendation systems are plagued with many challenges, such as adapting them to changes in user preferences and taste, and the effectiveness of recommendations made also determines the ability to retain and engage new users, as new user conversion to clients. This thesis proposes to use an adaptive hybrid collaborative approach to making recommendations to users. Four algorithms are combined: the Alternate Least Squares (ALS), KMeans clustering, Latent Dirichlet Allocation (LDA) and KMeans streaming. The recommender engine developed is in itself a multi-hybrid system as it not only combines four (4) algorithms but also combines the collaborative technique and content-based techniques of making a recommendation. Thus, the approach adopted can be used on datasets that contain rating information, textual descriptions or both. Three servers are leveraged in the implementation, consisting of the Scala server, PHP and Angular JS server and the MySQL database server for the storage of the results from the recommender engine. Various industry-standard metrics are adopted for the individual algorithms in addition to their computational times. These metrics include Root Mean Square Error(RMSE) for the ALS, Within Cluster Sum of Squares(WCSS) for KMeans, Log Perplexity and Log-Likelihood in the LDA. The memory estimates footprints and computational time on retraining the model are recorded for the KMeans streaming. The recommender engine is tested primarily on the 100K and 1M movieLens datasets and some portions of the 20M dataset are used. The implementation is compared with benchmark recommender algorithms via GitHub and existing offline implementations. In terms of retraining, the Adaptive Hybrid Collaborative Filtering Recommendation System(AHCF) developed improves a recommendation’s computational time concerning the offline model by 50%. The AHCF has an accuracy measure of 0.88-3.0 on RMSE values for the chosen datasets on increasing rank but less than 8 for 5 other datasets adopted. The other datasets range from restaurant datasets, anime, dating datasets, books and e-commerce. These results are taken for the 1M and 100K datasets. The unique contributions made in this research include combining multiple algorithms into one recommender engine that leverages textual and rating information at the same time. Improvements in computational efficiency as against offline models that are designed for a real-time update of recommendations by half on retraining. The generic nature of the algorithm also makes it useful to be used in many domains that leverage informative text and rating information. The model is also open source and available to all users. In a nutshell, the research embraces the efficiency of updating user preferences in real-time and making personalized recommendations by adapting to user preferences over short time intervals.
  • Thumbnail Image
    Item
    Improving Software Defect Prediction Using Cluster Under sampling.
    (University of Ghana, 2014-07) Agebure, M.P.; Sowah,R.A.; Amanquah, N.; University of Ghana, College of Basic and Applied Sciences, School of Engineering, Department of Computer Engineering
    Adequately learning and classifying datasets that are highly unbalanced has become one of the most challenging task in Data Mining and Machine Learning disciplines. Most datasets are adversely affected by the class imbalance problem due to the limited occurrence of positive examples. This phenomenon adversely affect the ability of classification algorithms to adequately learn from these data to correctly classify positive examples in new datasets. Data sampling techniques presented in Data Mining and Machine Learning literature are often used to manipulate the training data in order to minimize the level of imbalance prior to training classification models. This study presents an undersampling technique that has the capability of further improving the performance of classification algorithms when learning from imbalance datasets. The technique targets the removal of potential problematic instances from the majority class in the course of undersampling. The proposed technique uses Tomek links to detect and remove noisy/inconsistent instances, and data clustering to detect and remove outliers and redundant instances from the majority class. The proposed technique is implemented in Java within the framework of the WEKA machine learning tool. The performance of the proposed sampling technique has been evaluated with WEKA machine learning tool using C4.5 and OneR classification algorithms. Sixteen datasets with varying degrees of imbalance are used. The performance of the models when CUST is used are compared to RUS, ROS, CBU, SMOTE, OSS, and when no sampling is performed prior to training (NONE). The results of CUST are encouraging as compared to the other techniques particularly in datasets that have less than 2% minority instances and larger quantities of repeated instances. The experimental results using AUC and G-Mean showed that CUST resulted in higher performance in most of the datasets than the other methods. The average performance of the classification algorithms across the datasets for each technique also showed that CUST resulted in the highest average performance in all test cases. Statistical comparison of the mean performance also revealed that CUST performed statistically better than ROS, SMOTE, OSS and NONE in all test cases. CUST however, performed statistically the same as RUS and CBU, but with a higher mean performance. The results also confirmed that CUST is a viable alternative to the already existing sampling techniques particularly when the datasets are highly unbalanced with larger quantities of repeated, noisy instances and outliers.
  • Thumbnail Image
    Item
    Design and Development of Power Distribution Network Fault Data Collection Device, Fault Detection, Location and Classification Using Machine Learning Algorithms
    (University of Ghana, 2016-07) Dzabeng, N.A; Sowah, R.A; University of Ghana, College of Basic and Applied Sciences, School of Engineering, Department of Computer Engineering
    The detection and location of faults on power transmission lines is essential to the protection and maintenance of a power system. Most methods of fault detection and location rely on measurements of electrical quantities provided by current and voltage transformers. In this work, current sensors and voltage sensors were used in the prototyped model of the data collection device. Training data were collected by taking into consideration variables of a simulation situation like fault type, sensor location on the node, short circuit and open circuit faults were also analyzed. The test data were analyzed using three machine learning classifiers namely: K- Nearest Neighbor (KNN), Decision Tree and Support Vector Machines (SVM). Simulation result indicates that the anticipated method for fault classification gives high accuracy (99.5%) and least fault distance estimation error (0.5%) for all discussed cases. In order to verify the accuracy of the proposed method, a comparison is carried out with decision tree (DT), KNN and SVM. Separate investigation was also carried out with testing the system by varying the load at the range of 0%- 100%. It is observed from the test results of the network model that, the fault detection, location and classification gives a high accuracy with machine learning decision tree giving a quick training time response of 0.000999928 seconds.
  • Thumbnail Image
    Item
    Buoyant Gaussian Plume Modeling of atmospheric Dispersion of Cs-137 Emitted from a Nuclear Reactor
    (University of Ghana, 2013-07) Djangmah, J.A.
    A two dimensional Advection-Diffusion equation was developed to model atmospheric dispersion of Cs-137 in order to visualize the phenomenon of the transport ofCs-137 in the atmosphere. The study was also aimed at analyzing the activity concentration of Cs-137 and examined the mathematical approaches for determining the mass concentration of Cs-137 released to the atmosphere. The Buoyant Gaussian model was examined and a finite difference scheme was developed to solve the partial differential equations of radionuclide pollution transport in the atmosphere. The model program code for Gaussian Plume was implemented by using the MATLAB interface contouring software. The result outputs showed that like other parameters, depth has significant influence on the dispersion of pollutants; the highest and least activity concentrations of Cs-137 dispersion were found to be 0.4686 𝑚𝑔/𝑚3and 0.0002 𝑚𝑔/𝑚3 at 2.5 𝑚/𝑠 respectively. The maximum distance that was covered by Cesium-137 from the source downwind was 1621m at 5m/s. In the event of an accidental atmospheric release of radionuclides from a nuclear power plant, accurate forecasting of the dispersion and the activity concentrations of radionuclides are required for the preparation of adequate countermeasures and evacuation.
  • Thumbnail Image
    Item
    Simulation of Groundwater Flow in the Voltaian (Around Tamale) Using Carbon-14
    (University of Ghana, 2013-07) Nti-Agyemang, I.; Akiti, T.T.; Adomako, D.; University of Ghana, College of Basic and Applied Sciences, School of Engineering, Department of Computer Engineering
    Studying the process of groundwater flow in subsurface systems using numerical simulation has been widely practiced. The purpose of this study was to establish a 2D groundwater flow model for evaluating groundwater resources of the Voltaian Basin (around Tamale) in the Northern Region of Ghana. To understand the rate of abstraction of groundwater in the study area, a finite-element, steady-state groundwater flow model was used to simulate groundwater flow in the aquifer. COMSOL Multiphysics’ (FEMLAB) Earth Science Module (ESM) package which is finite element analysis and solver software was used. The radioisotope used in the study was Carbon-14. Three wells were sampled for Carbon-14 concentration and used for the model verification, based on elevation. From the results, groundwater in the study area moves generally from higher to lower hydraulic head along paths perpendicular to the equipotential lines. The groundwater flow paths in the aquifer in the study area indicated that flow is predominantly regional. There was a regional groundwater flow from Kanshegu to Nawuni. Kanshegu appears to be recharge area and Nawuni as discharge area. The flow rate obtained using Carbon-14 date was 2.86×10-7 m/s. The overall flow rate obtained from the model simulations was 2.66×10-7 m/s with an error margin of 6%.
  • Thumbnail Image
    Item
    Rotational Energy Harvesting to Prolong Flight Duration of Quadcopters
    (University of Ghana, 2014-07) Acquah, A.M.; Sowah, A.R.; Amanquah, N.
    This thesis presents a rotational energy harvester using brushless direct current (BLDC) generator to harvest ambient energy for quadcopter in order to prolong it flight duration. Quadcopters also known as drones are developed with the intention of operating in conditions where the presence of an on-board human pilot is either too risky or unnecessary, as such they are broadly referred to as UAV (Unmanned Aerial Vehicles). For a drone its endurance is essential in order to achieve operational goals, because most electrically powered drones have a limitation on size and mass, due to this they cannot carry a large mass of on-board energy thereby having short flight time. Quadcopters have a lot benefits such as large amount of controllability, hovering and manoeuvrability, because of this they are suitable for both indoor and outdoor applications such as scientific research, security surveillance and reconnaissance. BLDC generators are coupled with the propellers of the quadcopter to transfer kinetic= energy from the propellers to the generator. Taking into consideration the power requirement of quadcopter, the output of the generator is amplified using DC-DC boost and regulated to power and charge the on-board battery. The BLDC generator was simulated in MATLAB/SIMULINK, monitor and analyse the output of the generator. A final prototype of the rotational energy harvesting system was built and this comprised a quadcopter, power management system and a charging system. Results from the test conducted on the system produced output power levels of 4.98W at a source rotation speed of 5400RPM, which is use to augment the primary power supply. In all about 30% more energy was harvested from 4 micro-generators connected in parallel. This translates to about 10 minutes increase in flight endurance, thus a gain of about 50% in flight duration.
  • Thumbnail Image
    Item
    Network Infrastructure to Support Research and Education Network in Ghana
    (University of Ghana, 2007-01) Barfi-Adomako, O.; Tadayoni, R.; Buron, J.; University of Ghana, College of Basic and Applied Sciences, School of Engineering, Department of Computer Engineering
    This project examines and evaluates the campus-wide networks of five selected universities and two research institutions in Ghana. The nationwide backbone that would be needed to run these networks is also covered. Three of such backbones are identified as VRA Integrated fiber-optic backbone, New National backbone and Ghana Telecom country-wide networks. The Dwivedi and Wagner traffic model provides estimates of traffic demands based on published data on Ghana and selected educational nodes using 2005 as a reference year. Further traffic projections are given from 2005 to 2010 using results from the traffic model. The separation o f the traffic into voice, transaction data and internet traffic shows that, the Internet traffic will constitute about 45%, Voice (50%) and transaction data (5%) of the total traffic demand in Ghana by the end of the period under review. Dimensioning of the traffic using WDM Guru Software gives estimates of network cost and the equipment required to build the network. This enables different network topologies for Research and Education Networks in Ghana to be evaluated under various network protection schemes. The dimension results indicates that a pair of fiber cable would be required to build the backbone when Wavelength Division Multiplexing (WDM) and grooming techniques are deployed in the design of the network. The cost model, ownership options and organization structure suitable for the design and the implementation of the network in Ghana are covered in the report.