This repository contains code developed during a university lecture in machine learning for optimizing parameters of a neural network using grid and local search techniques. The implementation utilizes PyTorch for neural network modeling and optimization. Sample datasets for experimentation can be found in the data directory, while the preprocessing and parameter optimization procedures are provided in Python notebooks located in the src folder. Results obtained from the notebooks, including output logs and plots depicting training performance, are stored in the results folder.
The objective of this project is to optimize the parameters of a neural network model to achieve better performance on given datasets. The optimization process involves exploring different combinations of hyperparameters using both grid search and local search techniques. The implemented code allows for efficient experimentation and evaluation of various neural network configurations.
- data: Contains sample datasets for training and testing the neural network models.
- src: Contains Python notebooks implementing preprocessing steps, neural network modeling, and parameter optimization using grid and local search.
- results: Stores output logs, performance metrics, and visualization plots generated during the parameter optimization process.