neural-network-training

This is the code for Addressing Class Imbalance in Federated Learning (AAAI-2021).

Train neural audio and latent diffusion models (e.g. RAVE, vschaos2, AFTER) on Kaggle

Understanding the effects of data parallelism and sparsity on neural network training

Neural Network Training Fingerprint (NNTF) is a visualization approach to analyze the training process of any neural network performing classification.

Conjunto de ferramentas para lidar com treinamentos de redes neurais artificiais

Neural network for recognizing handwritten numbers

🧠 Automated neural network ablation studies using LLM agents and LangGraph. Systematically remove components, test performance, and gain insights into architecture importance through an intelligent multi-agent workflow.

An Open Source Machine Learning Framework for Everyone

Real-time Object Detection and Neural Network Training Visualization

Modifies a neural network's hyperparameters, activation functions, cost functions, and regularization methods to improve training performance and generalization.

This notebook provides source code of my work for my Master's Thesis.

Drift-Bounded Spectral Updates for Deep Local Learning

Replication code for the "Instance Space Analysis of Neural Network Training as Black-box Optimisation," comparing the CORNN benchmark suite with BBOB large-scale across dimensions 41, 261, and 481. Accepted to: ACM TELO (2026).

Simple implementation of a 3 layer MLP network using numpy.

Modifies a neural network's hyperparameters, activation functions, cost functions, and regularization methods to improve training performance and generalization.

This repository documents the process of finding the optimal learning rate for deep neural networks

Three optimization problem domains are created and applied to the randomized hill climbing, simulated annealing, genetic and MIMIC randomized optimization algorithms. Also, a neural network implementation is reimplemented using randomized optimization algorithms from the mlrose_hiive Python library.

A memory-efficient, gradient-free zeroth-order (derivative-free) optimizer designed to solve the "Curse of Dimensionality" in Black-Box optimization and memory-constrained Machine Learning. It provides an O(log D) gradient estimation approach that can successfully train Neural Networks without ever calculating analytical derivatives or Backprop