University Projects Repository

🎓 Showcase of my university projects for my CV 🚀

All projects' reports are in Greek because all subjects were taught in Greek language

📚 Table of Contents

1. Software Engineering
   • Earthquake Information Website with API
2. Microprocessors and Applications
   • Number conversion in ARM Processor
3. Integrated Circuits
   • Tic Tac Toe Game using SystemVerilog
4. Databases
   • Grading System Website with MySQL Database
5. Computer Intelligence
   • Knapsack Problem using Genetic Algorithm
   • Function Approximation using Neural Networks (MLP)
   • Fuzzy Cruise Control System
6. Pattern Recognition
   • Bayesian Classification & Decision Theory
   • Non-Parametric Estimation, SVM & Dimensionality Reduction (PCA/LDA)
   • Clustering, Feature Selection (PCA) & Multidimensional Scaling (MDS)
   • Final Term Project: Face Mask Detection & Chemical Binding Prediction
7. Algorithms and Complexity
   • RNA Secondary Structure
   • K-Center Problem Implementations (Pending)
8. Digital Signal Processing
   • Wiener Filter Implementation

Software Engineering

🚀 Earthquake Information Website with API

• Project Name: Earthquake Information Website with API
• Description: Briefly describe the project and its purpose. Include any relevant details such as programming languages used, frameworks, and technologies involved.
• Folder: [Link to project folder]

Microprocessors and Applications

🚀 Number conversion in ARM Processor

• Project Name: Number conversion in ARM Processor
• Description: Briefly describe the project and its purpose. Include any relevant details such as programming languages used, frameworks, and technologies involved.
• Folder: [Link to project folder]

Integrated Circuits

🚀 Tic Tac Toe Game using SystemVerilog

• Project Name: Hardware-Based Tic-Tac-Toe Game with AI Move Generator
• Description: A structural SystemVerilog implementation of a Tic-Tac-Toe game engineered for FPGA deployment. Features concurrent turn-management state machines, automated end-game win/tie evaluation logic, physical button debounce interfaces, and a specialized combinational AI Move Generator that dynamically computes optimal winning, blocking, or fallback moves.
• Folder: Integrated_Circuits

Databases

🚀 Grading System Website with MySQL Database

• Project Name: UniSystem - University Administration & Grading Information System
• Description: A comprehensive web-based management portal matching a MySQL relational database with a PHP/Bootstrap frontend. Features cryptographic session management, role-based access for students/professors/secretariat, dynamic DataTables filtering, and automated GPA/ECTS calculations.
• Folder: Databases

Computer Intelligence

🚀 Knapsack Problem using Genetic Algorithm

• Description: Implementation of a Genetic Algorithm (GA) in MATLAB to solve and maximize the total value of a 100-item 0-1 Knapsack Problem under capacity constraints. Applied critical performance optimization techniques (fitness score caching, removing redundant evaluation/import calls), drastically reducing execution runtime from 50-60 minutes down to just 3 – 4.5 seconds.
• Folder: Computer_Intelligence/Project1_Genetic_Algorithms

🚀 Function Approximation using Neural Networks (MLP)

• Description: Training a Multi-layer Perceptron (MLP) Feedforward Neural Network with 4 hidden layers and a [42, 28, 13, 7, 1] topology in MATLAB to approximate a non-linear mathematical function of two variables. Utilized Gradient Descent (traingd) combined with Early Stopping to successfully prevent premature overfitting.
• Folder: Computer_Intelligence/Project_2_Neural_Networks

🚀 Fuzzy Cruise Control System

• Description: Design and simulation of a Mamandani Fuzzy Inference System (FIS) in MATLAB to automatically control and stabilize a vehicle's cruising speed at 120 km/h. Modeled using tailored membership functions (Gaussian profiles combined with narrow triangular shapes (trimf)) for highly precise acceleration handling.
• Folder: Computer_Intelligence/Project_3_Fuzzy_Logic

Pattern Recognition

🚀 Bayesian Classification & Decision Theory

• Description: A series of 5 core tasks developed in Python (Jupyter Notebooks) covering: Bayesian email spam classification, mathematical definition of decision regions and experimental evaluation of minimal Bayes Risk, 3-class Multivariate Gaussian classifiers, theoretical and graphic worst-case error bound analysis, and Recursive Bayesian Parameter Estimation handled via Beta Distribution modeling.
• Folder: Pattern_Recognition/Project_1_Bayesian_Theory

🚀 Non-Parametric Estimation, SVM & Dimensionality Reduction (PCA/LDA)

• Description: Implementation of Non-Parametric Probability Density Estimation utilizing Parzen Windows (Kernel Density Estimation) to study the smoothing factor $h_N$. Training Linear and RBF Support Vector Machines (SVM) to evaluate and plot decision boundaries and Support Vectors. Dimensionality reduction and multi-feature evaluation executed over the Wine Dataset using Principal Component Analysis (PCA) and Fisher's Linear Discriminant Analysis (LDA), benchmarked via 10-fold Cross-Validation and Confusion Matrices.
• Folder: Pattern_Recognition/Project_2_Advanced_Classification

🚀 Clustering, Feature Selection (PCA) & Multidimensional Scaling (MDS)

• Description: Unsupervised learning pipeline evaluated on the morphological Seeds dataset using K-Means and Hierarchical Agglomerative Clustering across Euclidean and Cosine distances, benchmarked via Rand Index and Silhouette Coefficients. Features were optimized using PCA to preserve variance targets ($90%$ & $95%$). Includes coordinate map reconstruction from raw air distance matrices via Classical Multidimensional Scaling (MDS) in 2D and 3D spaces to study earth curvature constraints.
• Folder: Pattern_Recognition/Project_3_Clustering_MDS

🚀 Final Term Project: Face Mask Detection & Chemical Binding Prediction

• Description: Two-part advanced term project. Part 1 develops an automated Image Classification pipeline using custom Deep Convolutional Neural Networks (CNNs) and Transfer Learning via ResNet-18 for face mask compliance detection, optimized using fine-tuned classification boundaries for edge-case sorting. Part 2 addresses pharmacological drug discovery by applying PCA feature compaction and automated Bayesian search to tune deep MLPs, SVM regressions, and soft/hard Ensemble Voting models predicting molecular binding affinities, validated via AUC-ROC evaluation.
• Folder: Pattern_Recognition/Final_Term_Project

Algorithms and Complexity

🚀 RNA Secondary Structure

• Project Name: RNA Secondary Structure
• Description: Given an RNA sequence, calculate the RNA Secondary structure with the most bonds possible. Dynamic programming was used for the important function of the project. The project is implemented in python using a Jupyter Notebook.
• Folder: RNA_Secondary_Structure

🚀 Project 6.2

• Project Name: [Project Name]
• Description: Briefly describe the project and its purpose. Include any relevant details such as programming languages used, frameworks, and technologies involved.
• Folder: [Link to project folder]

Digital Signal Processing

🚀 Wiener Filter Implementation

• Project Name: Audio Denoising & Linear Prediction using Wiener Filters
• Description: GMATLAB framework implementing optimal Wiener Filtering strategies for audio restoration. Features global and short-time frame-based processing (Hanning windowed Overlap-Add), adaptive noise covariance estimation from unvoiced silent intervals, and multi-step forward Linear Prediction modeling evaluated via SNR benchmarking.
• Folder: DSP

📝 License

This repository is licensed under the MIT License.

📧 Contact

If you have any questions or would like to get in touch, feel free to reach out to me at kon.iatrou06@gmail.com.

_{Built by Konstantinos Iatrou}