tommaso-ferracci
Aspiring Data Scientist, currently studying Statistics at Imperial College London.
Imperial College LondonLondon - United Kingdom
Pinned Repositories
CTGAN
Conditional GAN for generating synthetic tabular data.
Deep_Learning_Energy_Reconstruction_JUNO
BSc Thesis. Energy reconstruction of neutrino interactions at the JUNO experiment via BDT (xgboost), FCDNN and 1DCNN (pytorch).
GPT_Table_To_LaTeX
Simple Streamlit app which allows for the conversion of a .csv table to LaTeX code using GPT-3.5-turbo.
Heat_Equation_2D
Solution to the 2D heat equation in a square simulation cell by using both explicit and implicit Euler methods. The instability of the former is discussed.
Metropolis_Lennard_Jones
Metropolis Monte Carlo simulation of a system of N particles in a Lennard Jones potential obeying classical equations of motion under periodic boundary conditions.
Parallel_Coordinates_Plot
Python package to create parallel coordinates plots of the hyperparameter search supporting most of the modern machine learning and deep learning frameworks.
Prigozhin_Audio_Files
Analysis of Yevgeny Prigozhin's audio files on Telegram using the GPT-3.5-turbo API. Part of a data journalism coursework at Imperial College London.
Reinforcement_Learning_Lunar_Lander
Solution to the lunar lander environment from OpenAI's gym making use of expected SARSA with neural networks.
Resonance_Curve_Regression_Analysis
Non-linear fit of the resonance curve in a RLC circuit. The estimates of parameter errors given by the minimizer are verified through chi-square contour plots and toy-datasets simulations.
Schrodinger_Equation
Numerical solutions to the 1D and 2D Schrodinger equation, both stationary and time-dependent. These include the Numerov propagator, the Crank-Nicolson method and an implicit method using scipy's sparse matrices.
tommaso-ferracci's Repositories
tommaso-ferracci/Deep_Learning_Energy_Reconstruction_JUNO
BSc Thesis. Energy reconstruction of neutrino interactions at the JUNO experiment via BDT (xgboost), FCDNN and 1DCNN (pytorch).
tommaso-ferracci/GPT_Table_To_LaTeX
Simple Streamlit app which allows for the conversion of a .csv table to LaTeX code using GPT-3.5-turbo.
tommaso-ferracci/Heat_Equation_2D
Solution to the 2D heat equation in a square simulation cell by using both explicit and implicit Euler methods. The instability of the former is discussed.
tommaso-ferracci/Parallel_Coordinates_Plot
Python package to create parallel coordinates plots of the hyperparameter search supporting most of the modern machine learning and deep learning frameworks.
tommaso-ferracci/Resonance_Curve_Regression_Analysis
Non-linear fit of the resonance curve in a RLC circuit. The estimates of parameter errors given by the minimizer are verified through chi-square contour plots and toy-datasets simulations.
tommaso-ferracci/Schrodinger_Equation
Numerical solutions to the 1D and 2D Schrodinger equation, both stationary and time-dependent. These include the Numerov propagator, the Crank-Nicolson method and an implicit method using scipy's sparse matrices.
tommaso-ferracci/CTGAN
Conditional GAN for generating synthetic tabular data.
tommaso-ferracci/Metropolis_Lennard_Jones
Metropolis Monte Carlo simulation of a system of N particles in a Lennard Jones potential obeying classical equations of motion under periodic boundary conditions.
tommaso-ferracci/Prigozhin_Audio_Files
Analysis of Yevgeny Prigozhin's audio files on Telegram using the GPT-3.5-turbo API. Part of a data journalism coursework at Imperial College London.
tommaso-ferracci/Reinforcement_Learning_Lunar_Lander
Solution to the lunar lander environment from OpenAI's gym making use of expected SARSA with neural networks.
tommaso-ferracci/Targeted_Augmentation_Amex
MSc thesis - Imperial College London. Identification of the hardest datapoints via KNN Shapleys + synthetic data generation on large credit default dataset provided by American Express.
tommaso-ferracci/SDV
Synthetic data generation for tabular data
tommaso-ferracci/tommaso-ferracci
My personal GitHub README.