AAU Institutional Repository (AAU-ETD)
Addis Ababa University Institutional repository is an open access repository that collects,preserves, and disseminates scholarly outputs of the university. AAU-ETD archives' collection of master's theses, doctoral dissertations and preprints showcase the wide range of academic research undertaken by AAU students over the course of the University's long history.
How to Submit Your Work
The repository contains scholarly work, both unpublished and published, by current or former AAU faculty, staff, and students, including Works by AAU students as part of their masters, doctoral, or post-doctoral research
- All AAU faculty, staff, and students are invited to submit their work to the repository. Please contact the library at your college.
You may contact digirep@aau.edu.et.with any questions about the repository
Colleges,Institutes in AAU-ETD
Select a college,institute to browse its collections.
Recent Submissions
Deep Learning Based SIMBox Fraud Detection using CDR data: A case of Safaricom Ethiopia
(Addis Ababa University, 2024-06) Fikirte Endalew; Elefelious Getachew (PhD)
The telecommunications industry is a critical component of modern society, facilitating communication and data exchange across individuals and businesses globally. However, this interconnectedness also presents vulnerabilities that malicious actors exploit. In the telecom sector, fraud usually refers to deliberate misuse of voice and data networks as well as service theft. One of the most difficult problems telecom organizations worldwide have is SIMBox fraud. A SIMBox fraud diverts international calls to a cellular device through the internet via a device called a SIMBox, routing telecom services to local networks into the network as local services, using hundreds of low-cost or even unpaid SIM cards, which are often obtained with forged identities. Ethiopia’s distinctive ethnolinguistic, cultural, and socioeconomic landscape significantly shapes its Call Detail Record (CDR) data. To effectively detect SIM Box fraud within this context, it is imperative to develop fraud detection models that are specifically tailored to the Ethiopian telecommunications environment. Detecting fraud activities becomes increasingly challenging as the number of subscribers and CDR log volumes and velocities increase. In order to identify telecom fraud via data mining techniques, deep learning techniques have become more and more popular in the telecom sector and other domains in recent years. While Machine Learning algorithms demonstrate effectiveness in detection, a fundamental challenge lies in balancing speed with accuracy. This challenge requires a careful balance between the two, as optimizing one metric often compromises the other. Telecom operators are facing financial losses due to SIM box fraud. Early detection of these fraudulent activities is critical to minimize revenue leakage. Therefore, evaluating the effectiveness of various fraud detection systems is essential to ensure a swift response. In this thesis, a CRISP-DM based methodology is followed to collect, discover, preprocess and model as well as evaluate CDR based SIMBox fraud detection for Safaricom Ethiopia. BERT, MLP, LSTM and classic rRNN deep learning models are implemented with evaluation. The results show that the rRNN algorithm with GRU architecture showed the highest accuracy of 99.7% followed by LSTM, BERT and MLP at 99.1%,98.6% and 96.7% of accuracy respectively.
Parametric Study of Reinforced Concrete Beams Strengthened in Flexure by Fiber Reinforced Polymer
(Addis Ababa University, 2025-05) Kaleab Teshome; Abrham Gebre (PhD)
This study examines the effect of parameters on damaged Reinforced Concrete (RC) beams by using Finite Element Analysis (FEA) method. A software verification was conducted for the ABAQUS software by using an experimental result that is found in a literature review. The result shows that the FEA method gives approximately similar results to the experimental results. A Latin Hypercubic Sampling (LHS) method was used to create 32 samples that have 7 parameters. The FEA software was used to conduct a non-linear analysis for 32 samples of damaged RC beams that are strengthened with Carbon fiber Reinforced Polymer (CFRP) on the sides. The non-linear FEA results indicated that the stiffness of strengthened RC beams increased compared to un-strengthened beams. An equation that can be used to determine the extent of flexural strengthening that can be applied on the side of a damaged RC beams has been formulated by using regression analysis. According to a sensitivity analysis conducted to determine which parameter affects most amongst the parameters of degree of damage, ratio of length of CFRP to length of beam (LCFRP/Lbeam), ratio of width of CFRP to depth of beam (bCFRP/hbeam), number of layers of CFRP, compressive strength of concrete, yield strength of reinforcement and tensile reinforcement ratio (ρ) in a strengthening method, it is shown that the tensile reinforcement ratio (ρ) and the ratio of the length of the CFRP to the length of the beam (LCFRPP/Lbeam) have showed the highest direct effects on the load carrying capacity of strengthened RC beams. While the degree of damage showed an inverse effect. A verification of the equation was conducted using both strengthened and un-strengthened RC beams. The results showed that the equation should only be used for strengthened RC beams. The equation was also compared with other analytical equations provided by researchers and codes such as ACI 440.2R-08, Bai Y-L et al., Stephen Lee and CEB-FIP technical report [1, 2, 3, 4]. The comparison shows that the FIB technical report and the ACI440.2R-08 are in good agreement with the proposed equation. The comparison also showed that the proposed equation is conservative than the researches conducted by Bai Y-L et al. and Stephen Lee.
Experimental Study on PPC and Citric Acid Influence on Compressive Strength of Reinforced Concrete
(Addis Ababa University, 2025-06) Abebech Haileselassie; Sintayehu Nibert (PhD)
The aim of this experimental study is to investigate the combined effect of partially replacing ordinary Portland cement (OPC) with Portland Pozzolana Cement (PPC) and incorporating citric acid by weight of PPC as chemical admixture on the compressive strength performance of reinforced concrete. PPC, known for its environmental benefits and long-term strength development, is evaluated alongside citric acid, which acts as a set retarder and potential performance enhancer. The experimental work includes tests on setting time, workability, and compressive strength for concrete mixes incorporating varying dosages of citric acid with PPC. The experimental research involved creating concrete samples specifically for C-30 concrete grades, a common choice for reinforced concrete projects, having water cement ratio of 0.50. Concrete samples are created using OPC, PPC and also PPC mixed with citric acid & slump in the range of 50 – 100mm. Citric acids was added to retard the setting time, to enhance the workability & ultimate compressive strength of concrete. Setting time, Slump & compressive strength tests for 0.1%, 0.2%, 0.3% & 0.4 % additions of citric acid by weight of Portland Pozzolanic Cement were performed. Cost comparison was also done between concrete mixes made with OPC and PPC with Citric acid. The laboratory results assure that the optimum dosage of citric acid is 0.3% by weight of PPC cement. This dosage causes a delay in initial and final setting time of 1hr 50 min & 3hr 10 min, respectively; 15% water reduction; a comparable compressive strength result with OPC mix and meets the requirements listed in ACI 301; and an increase in slump length of 55% relative to reference concrete. The laboratory test result shows that as compared with the reference mix, satisfactory setting time, slump & compressive strength result was obtained from concrete mix made with PPC in combination with Citric acid. Moreover, the combination of PPC and citric acid demonstrates promising potential for sustainable and performance-optimized reinforced concrete. This study contributes valuable insights for engineers and researchers seeking to enhance the performance and sustainability of concrete construction materials.
BWAF-Net: Enhanced Human Promoter Identification via Biologically Weighted Attention Fusion of Transformer and Graph Attention Networks
(Addis Ababa University, 2025-10) Zemedkun Abebe; Adane Letta (PhD)
The identification of gene promoter regions is crucial for understanding transcriptional regulation, yet computational methods often struggle to effectively integrate the diverse biological signals involved. Existing approaches typically focus on a single data modality, such as the DNA sequence, or employ simple fusion techniques that fail to leverage explicit biological knowledge. To address these limitations, we present BWAF-Net, a novel multi-modal deep learning framework for the identification of human promoters. BWAF-Net integrates three data streams: DNA sequences processed by a Transformer to capture long-range dependencies; gene regulatory context from 36 tissue-specific networks modeled by a Graph Attention Network (GAT); and explicit domain knowledge in the form of 11 quantified biological motif counts (priors). The framework’s central innovation is the Biologically Weighted Attention Fusion (BWAF) layer, which uses the biological priors to learn dynamic attention weights that modulate the fusion of the sequence and network representations. Evaluated on a balanced dataset of 40,056 human promoter and non-promoter sequences, BWAF-Net achieved outstanding performance, with 99.87% accuracy, 99.99% AUC-ROC, and 100% precision on the held-out test set. The proposed framework significantly outperformed a replicated state-of-the-art, sequence-only baseline as well as a series of ablated models. Our ablation studies confirm that naive feature concatenation is a suboptimal fusion strategy, validating the necessity of the intelligent BWAF mechanism. By providing a framework that is highly accurate, parameter-efficient, and interpretable, this work presents a significant advance in multi-modal AI for regulatory genomics.
Advancing Amharic Text Summarization with a Tailored Parameter-Efficient Fine-Tuning Technique
(Addis Ababa University, 2025-08) Dagim Melkie; Fantahun Bogale (PhD)
While recent progress in Large Language Models (LLMs) has revolutionized the field of Natural Language Processing (NLP), applying these models to low-resource languages such as Amharic presents considerable difficulties. Key obstacles include the scarcity of available data and the intensive computational cost associated with conventional finetuning methods. To overcome these issues, this thesis introduces a specialized parameterefficient fine-tuning (PEFT) framework developed specifically for Amharic text summarization. This new framework combines a dynamic low-rank adaptation component
(DyLoRA-Amharic) with an adaptive activation method (AdaptAmharic), which work together to improve the model’s flexibility and optimize its resource allocation during training. The methodology involves injecting these custom modules into the mT5-small encoder– decoder architecture, allowing dynamic adjustment of DyLoRA-Amharic ranks and AdaptAmharic activation levels based on gradient signals. A joint optimization objective incorporating regularization terms for both rank and activation was employed to manage model complexity and ensure training stability. Comparative experiments were conducted against standard PEFT LoRA and Houlsby Adapter baselines on a curated Amharic summarization dataset. Experimental results demonstrate that the proposed DyLoRA-Amharic and Adapt Amharic framework significantly out performs the baselines across ROUGE, BLEU, and BERT Score metrics, achieving the lowest evaluation loss. Specifically, it improved ROUGEL by 30.5% and BLEU by 52.4% over the strongest baseline. This superior performance validates the efficacy of a densely injected, dynamic, and regularized architecture, challenging the conventional emphasis on maximal sparsity in PEFT. While the framework utilizes a higher proportion of trainable parameters (13.42%) compared to the baselines, this trade-off is justified by the substantial performance gains. This research contributes to advancing PEFT methodologies for low-resource NLP, providing a robust and adaptable solution for Amharic text summarization. The findings lay a foundation for developing more efficient and effective LLMs for diverse and linguistically underrepresented communities