Machine Learning with Python (IBM - Coursera)

Welcome to my personal repository for the Machine Learning with Python course by IBM on Coursera. I created this space to organize my course materials, including notebooks, code examples, and additional resources, as I work through the exciting world of machine learning using Python.

Table of Contents

• Overview
• Course Modules
• Final Project
• What I've Learned
• Prerequisites
• Installation
• Usage
• Directory Structure
• Resources
• License
• Acknowledgments

Overview

I enrolled in this course to build a solid understanding of machine learning concepts and practical applications using Python. Through a series of lectures, hands-on assignments, and interactive Jupyter notebooks, I am learning about both supervised and unsupervised learning techniques. This repository serves as a living document of my journey, capturing my notes, experiments, and reflections as I progress through the material.

Course Modules

The course is divided into several modules, and here’s a quick overview of each:

• Introduction to Machine Learning:
  I received an overview of machine learning fundamentals, its applications, and the basic tools required for building models.

• Data Preparation and Analysis:
  This module covered techniques for cleaning, preprocessing, and analyzing data—crucial steps before feeding data into machine learning algorithms.

• Supervised Learning:
  I explored regression, classification, and various evaluation metrics to understand how models are built and validated.

• Unsupervised Learning:
  Here, I learned about clustering, dimensionality reduction, and pattern discovery, which are useful when working with unlabeled data.

• Practical Applications:
  This part of the course allowed me to apply what I learned to real-world problems through case studies and project-based learning.

Each module is organized in its own folder within this repository.

Final Project

As a capstone project for this course, I built a machine learning pipeline to predict whether it will rain today in Melbourne, Australia, using historical weather data. Here's a high-level summary of what I implemented:

• Data Source:
  Weather data from multiple Melbourne locations, cleaned and filtered for relevant entries.

• Feature Engineering:
  Mapped dates to seasons, renamed columns for clarity, and separated numerical and categorical features.

• Preprocessing Pipeline:
  Used ColumnTransformer to scale numeric features and one-hot encode categorical ones. Handled unknown categories and ensured the pipeline could be reused for both training and inference.

• Model Training & Tuning:
  Built and tuned a RandomForestClassifier and LogisticRegression using GridSearchCV with 5-fold cross-validation and stratified splits. Optimized for accuracy and balanced class weights due to class imbalance.

• Evaluation:
  Reported accuracy, precision, recall, and F1-score. Displayed confusion matrices using both matplotlib and seaborn for visual analysis. Identified top contributing features to model predictions using feature importance scores.

This project gave me valuable hands-on experience in handling imbalanced classification problems, designing preprocessing workflows, evaluating multiple models, and visualizing performance.

What I've Learned

So far, the course has helped me:

• Understand the core principles of machine learning and its various applications.
• Get hands-on experience with Python libraries like NumPy, pandas, matplotlib, and scikit-learn.
• Learn the importance of data cleaning and preparation before model training.
• Build, evaluate, and fine-tune predictive models.
• Appreciate the role of both supervised and unsupervised learning in solving different types of problems.
• Build end-to-end pipelines for real-world classification tasks.

This journey has laid a strong foundation for my continued exploration of machine learning.

Prerequisites

Before diving into this course, it's helpful to have:

• A basic understanding of Python programming.
• Familiarity with libraries such as NumPy, pandas, matplotlib, and scikit-learn.
• An environment capable of running Jupyter Notebooks (e.g., Anaconda or a virtual environment with the necessary packages).

Installation

To set up your environment, follow these steps:

1. Clone the Repository:

   git clone https://github.com/emsilver987/Machine-Learning-Python-IBM
   cd Python-IBM-Project

2. (Optional but recommended) Create a virtual environment:

   python -m venv venv
   source venv/bin/activate        # On Windows use: venv\Scripts\activate

3. Install the required packages:

   pip install -r requirements.txt

Contact

If you'd like to learn more about the project or have any feedback, feel free to connect with me on LinkedIn or reach out via email esilvert@purdue.edu