Astrophysicist | Data Scientist | Machine Learning Enthusiast
Welcome to my GitHub profile! Here, you'll find a collection of my work in data science, artificial intelligence, and astrophysics, reflecting my passion for leveraging cutting-edge technology to solve complex problems.
I am a graduate of the M.Sc. program in Astronomy and Data Science at Leiden University, with a B.Sc. in Physics from Aristotle University of Thessaloniki. Beyond academia, I am passionate about data science, machine and deep learning, reinforcement learning, and statistics. Many of my projects in these fields have been driven by curiosity and a desire to solve practical problems. I am always excited about open-source projects, collaborative learning, and applying AI to new challenges.
Predicting binary neutron star postmerger spectra using artificial neural networks
Dimitrios Pesios, Ioannis Koutalios, Dimitris Kugiumtzis, Nikolaos Stergioulas
Physical Review D, Sep 2024
DOI:
10.1103/PhysRevD.110.063008
arXiv:2405.09468
A selection of my repositories, showcasing work in data mining, machine learning, astrophysics, and more. Each project reflects my commitment to solving complex problems using cutting-edge tools and techniques.
AdaptiveOptics
Reinforcement learning applied to optimize adaptive optics systems in the absence of sensor data.
DeepLearning
Experiments with neural network architectures for various machine learning tasks. Also includes implementations of neural networks from scratch.
PredictFutureSales
Forecasts store sales in this Kaggle competition using machine learning algorithms like LightGBM and Random Forest.
RecommenderSystem
Multiple recommender systems in a challenge similar to the classic Netflix Prize.
AdventOfCode2024
Solutions to the Advent of Code 2024 challenges, using Python.
Titanic
Survival prediction on the Titanic using classical machine learning techniques.
Spaceship-Titanic
Experimenting with "scikit-learn" on the futuristic Spaceship Titanic dataset.
g2net_malta_hackaton
Distinguishing earthquake signals from ambient noise using classical machine learning models.
g2net_thessaloniki_hackaton
Working with real gravitational wave data to characterize the Signal-to-Noise Ratio, using machine learning techniques, and carefully tuning hyperparameters.
Matched-filtering
A tool for gravitational wave detection using matched-filtering algorithms.
Stellar Evolution
Simulating stellar life cycles and comparing the evolution of stars with different masses. The MESA code, written in Fortran, is used for this project.
Exoplanet Project
A project on planet evolution for different masses and compositions, using the MESA code.
Languages, Frameworks and Tools (Expert) Python Jupyter Git Linux OS LaTeX VS Code
|
Languages, Frameworks and Tools (Intermediate) SQL C/C++Mathematica  Matlab HTML CSS
|
Machine and Deep Learning PyTorch TensorFlow Keras Scikit-learnJAX |
Data Science Pandas Numpy MatplotlibScipy Seaborn Nbodykit |
I.T. and Networking Bash SSH Anaconda Docker GitHub
|
Soft Skills Problem Solving Critical Thinking Teamwork Communication Time Management Adaptability |