Machine Learning Portfolio

This repository contains a collection of applied machine learning projects completed using Python, Jupyter Notebook, and machine learning/deep learning frameworks. The projects demonstrate core concepts in data preprocessing, dimensionality reduction, supervised learning, neural networks, convolutional neural networks, classification modeling, prediction workflows, and basic model deployment.

Repository Overview

This portfolio includes multiple machine learning assignments and project files, including:

• Principal Component Analysis (PCA)
• Neural Networks
• Convolutional Neural Networks (CNNs)
• Classification model development
• Prediction output generation
• Flask-based machine learning web application
• Saved trained model files

Projects Included

1. Principal Component Analysis (PCA)

This project applies Principal Component Analysis to reduce feature dimensionality while preserving important information in the dataset. PCA is commonly used for exploratory data analysis, feature reduction, visualization, and improving model efficiency.

Key concepts demonstrated:

• Dimensionality reduction
• Feature transformation
• Explained variance
• Data visualization
• Python-based machine learning workflow

2. Neural Networks

This project focuses on building and evaluating neural network models for classification or prediction tasks. It demonstrates the use of deep learning concepts such as layers, activation functions, model training, and evaluation.

Key concepts demonstrated:

• Artificial neural networks
• Model training and validation
• Classification modeling
• Deep learning fundamentals
• Python machine learning libraries

3. Convolutional Neural Networks (CNNs)

This project explores convolutional neural networks, which are commonly used for image classification and computer vision tasks. The project demonstrates how CNN architectures can extract patterns from image or structured input data.

Key concepts demonstrated:

• Convolutional neural networks
• Deep learning model architecture
• Image or pattern recognition
• Model training
• Model evaluation

4. Machine Learning Web Application

This repository also includes a Flask-based machine learning application. The application uses user-provided inputs, preprocesses the data, loads a trained model, and returns a prediction. The Flask app includes routes for an about page and prediction pages, with a prediction output displayed through an HTML template.

Key concepts demonstrated:

• Flask web application development
• Model loading
• User input handling
• Prediction workflow
• HTML templates
• Machine learning deployment basics

Files in This Repository

Example files include:

• Assignment_1_Principal_Component_Analysis_(PCA).ipynb
• Assignment_2_Neural_Networks.ipynb
• Assignment_3_CNNs.ipynb
• Assignment 4.ipynb
• FinalAssignment_GilbertMorgan.ipynb
• classification_model.h5
• CNN_model.h5
• Gilbert_PCA_mushroom_model.h5
• application.py
• about.html
• predict.html

Tools and Technologies

• Python
• Jupyter Notebook
• Pandas
• NumPy
• Scikit-learn
• TensorFlow / Keras
• Flask
• HTML
• Machine Learning
• Deep Learning

Skills Demonstrated

• Data cleaning and preprocessing
• Exploratory data analysis
• Principal Component Analysis
• Model training and evaluation
• Neural network development
• CNN development
• Classification modeling
• Model saving and loading
• Web-based prediction deployment
• Research documentation and reproducible workflows

Purpose

The purpose of this repository is to document and showcase my applied machine learning coursework and project development. It reflects my ability to work with machine learning models from the data preparation stage through model training, evaluation, and basic deployment.

Author

Gilbert Morgan

Data Science Graduate Student
Machine Learning | Data Analysis | Python | Predictive Modeling