Applied Machine Learning (AML)

A collection of hands-on machine learning projects completed as part of the Applied Machine Learning course at Indiana University. These notebooks explore real-world datasets, classification models, and data preprocessing techniques.

---

Projects Overview

Project	Description
Insurance Data	Regression or classification models predicting insurance outcomes (details TBD based on data).
Sign MNIST	Handwritten digit recognition using the Sign MNIST dataset.
Fashion MNIST	Fashion item classification (e.g., T-shirt, shoe) using the Fashion MNIST dataset.

Project Highlights

Sign MNIST & Fashion MNIST

• Preprocessing: Rescale, normalize, and augment images.
• Modeling: Train and compare performance of:
  • Convolutional Neural Networks (CNNs)
  • Classic classifiers (Logistic Regression, SVM)
• Evaluation:
  • Calculate accuracy, precision, recall, and F1-score.
  • Visualize confusion matrices and training curves.

Insurance Data

• Exploratory Data Analysis (EDA): Identify missing values, outliers, and feature distributions.
• Modeling:
  • Tabular models like Random Forests, Gradient Boosted Trees, or Linear Regression.
• Feature Engineering:
  • Generate new features (e.g., ratios, binned categories) to improve model accuracy.

---

How to Use This Repository

1. Clone the repo:

   git clone https://github.com/git-avinashpawar/AML.git
   cd AML

2. Set up a virtual environment (recommended):

   python3 -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt

3. Launch a notebook:

   jupyter notebook
   #Open the desired .ipynb file (e.g., Sign_MNIST.ipynb) to explore modeling code, results, and visualizations.