readme: update project structure and abstract

Author: Favo02

README.md

@@ -1,28 +1,51 @@
-## AdaBoost
+# AdaBoost
 
-Project for [Statistical Methods for Machine Learning](https://cesa-bianchi.di.unimi.it/MSA/index_25-26.html) course at Università degli Studi di Milano - Master degree in Computer Science (a.y. 2025/26).
+Project for the [Statistical Methods for Machine Learning](https://cesa-bianchi.di.unimi.it/MSA/index_25-26.html) course at Università degli Studi di Milano - Master's degree in Computer Science (a.y. 2025/26).
+
+Compiled Report PDF: [report.pdf](./report/report.pdf)
 
 ## Abstract
 
-TODO
+The project implements the Adaptive Boosting algorithm (**AdaBoost**) from scratch using **decision stumps** (decision trees of depth 1) as weak learners.
+The aim of the study is to understand how boosting combines weak learners, focuses on hard examples, and minimizes the classification error over training rounds.
+
+Experiments are run on one real-world dataset (UCI Adult) and three synthetic datasets (linearly separable, XOR, concentric circles).
+AdaBoost is then compared against a single **decision tree** and **logistic regression**.
+Results show that AdaBoost performs very well on most datasets, but completely fails on the XOR dataset where simple decision stumps cannot beat random guessing.
 
 ## Project Structure
 
-```
-├── notebook.ipynb       # results and graphs
+```text
+├── 01_datasets.ipynb          # Dataset loading and preprocessing
+├── 02_hyperparameters.ipynb   # Hyperparameters tuning
+├── 03_adaboost.ipynb          # AdaBoost analysis
+├── 04_comparison.ipynb        # Baseline models comparison
 ├── src/
-│   ├── __init__.py
-│   ├── data.py          # dataset
-│   └── adaboost.py      # adaboost from scratch
-├── uv.lock              # project metadata
-└── pyproject.toml       # project metadata
+│   ├── data.py                # Data loading and generation
+│   ├── decision_stump.py      # Weighted decision stump weak learner
+│   ├── decision_tree.py       # Decision tree
+│   ├── logistic_regression.py # Logistic regression
+│   ├── adaboost.py            # AdaBoost ensemble class
+│   └── hyperparameters.py     # Helper functions for hyperparameter tuning
+├── report/
+│   ├── report.typ             # Source of the report
+│   └── report.pdf             # Compiled report PDF
+└── figs/                      # Plots and figures
 ```
 
-## Run
+## Setup
+
+Dependencies and virtual environments managed by [uv](https://github.com/astral-sh/uv).
+Formatting and linting using [ruff](https://github.com/astral-sh/ruff) and [ty](https://github.com/astral-sh/ty).
 
-Dependencies and venv managed by uv _(which is better than pip)_.
+Install dependencies:
 
 ```bash
 uv sync
+```
+
+Run the notebooks:
+
+```bash
 uv run jupyter lab
 ```