Marko Njegomir's Projects
Microservice AI Neural Network Tensorflowjs face (YOLO object) detection Two-factor deep learning (TensorFlow) authentication in Spring, Vue, and Flask with face verification based on logistic regression and cosine similarity in high dimensional vector space.
Minmax (minimax) algorithm with Alpha-Beta (𝛼−𝛽, ab) Pruning optimization for the Checkers (Draughts) game in Python 3. Same minimax algorithm works for chess, tic tac toe or similar games.
Curriculum learning in computer vision. CNN models are trained on the dataset, and then the dataset is split based on the difficulty.
Predmet: Nelinearno programiranje i evolutivni algoritmi Tema: Genetski algoritam, problem optimizacije kontinualnih funkcija Tri funkcije su: (Ackley, Griewank, Michalewicz )
Traveling salesman problem solved using genetic algorithm
GATor is a Graph Attention Network for object detection with relational reasoning
Graph Neural Network (GNN) for knowledge graph completion.
Graph Neural Network (GNN) Transformer for predicting drug-drug interactions (ddi, polypharmacy) from drug molecules.
Graph Neural Network (GNN) recommendation system made with Pytorch Geometric.
Experiments with transformer synopsis embeddings for GNN recommender system.
Python program for searching pdf text, ranking the results and exporting highlighted search results in pdf. Uses trie structure, stack, heap, page graph. Converts queries to postfix notation. Allows for logical expressions and phrases. Offers did you mean functionality.
Transformer implemented with graph attention network (GAT) layers from PyTorch Geometric
Human-computer interaction project for organizing events with C# WPF, Material design, Entity Framework Core.
Mastermind game coded in i386 assembly language. It uses only one 32 bit register to store the combination.
Integrated development environment for i386 assembly and C programming language
Restaurant ordering and management application. Fully tested with unit, integration and e2e Selenium tests. Spring-boot on backend, Angular on frontend.
Restaurant ordering and management application. Fully tested with unit, integration and e2e Selenium tests. Spring-boot on frontend, angular on backend.
A new type of sorting algorithm. Use large language model (llm like gpt, chat-gpt or others) to sort collections.
My solutions for Stanford University course CS224W: Machine Learning with Graphs Fall 2021 colabs (GNN, GAT, GraphSAGE, GCN)
Mastermind game with changed rules so that all possible combinations are calculated and points are adjusted accordingly. Known as "Skočko" in Serbia.
Language for the mastermind game in C with Flex and Bison. C++ program is used to generate zero-byte files, and to decode them. Make tool is used to create the compiler and run replays.
Expert rule based system for tracking inputs from neuralink chip and smart-watch in order to improve daily habits, measure sleep efficiency and improve sleep quality.
Microservice architecture for publishing papers. Custom implementation with Spring Boot, RabbitMQ, PostreSQL, MongoDB and RestTemplate
Microservice architecture for publishing scientific papers. Created with Silvera tool. Java, Spring-Boot, MongoDB, Kafka
Smart home with security. IoT. HTTPS (SSL/TLS), CSR, Certificates, Spring-boot, Angular, Drools, RabbitMQ, MongoDB, PostgreSQL.
Django and D3 Team project for SOIK (Design patterns and components)
ER model for Space Missions and sql scripts that create relational database based on the model and PLSQL scripts that generate reports