Data Science & Machine Learning Portfolio

A curated collection of 11 data science & machine learning projects demonstrating practical expertise across the full ML lifecycle: from data exploration and feature engineering to model deployment and evaluation.

Created by Kwami Alfred Jordal BIAM -- Data Scientist & AI Engineer.

Overview

This portfolio showcases hands-on experience with the most in-demand data science and AI technologies. Each project is an end-to-end implementation covering problem formulation, data preparation, exploratory analysis, modeling, evaluation, and interpretation.

What makes this portfolio stand out:

Breadth: Covers NLP, deep learning, time series, dimensionality reduction, vector search, and LLMs
Depth: Each project includes rigorous statistical analysis, model diagnostics, and business-oriented interpretations
Production awareness: Projects use industry-standard tools (Pinecone, HuggingFace, TensorFlow) and follow software engineering best practices
Reproducibility: Every notebook is self-contained with clear markdown explanations and documented results

Projects

#	Domain	Project	Techniques	Highlights
1	Natural Language Processing	Fake News Detection	TF-IDF, BoW, Logistic Regression, Naive Bayes, Linear SVM, LDA Topic Modeling	99.3% accuracy with Linear SVM
2	Deep Learning	Customer Churn Prediction	TensorFlow/Keras, Neural Networks, Adam Optimizer	End-to-end TF pipeline
3	Dimensionality Reduction	Music Genre Classification with PCA	PCA, Logistic Regression, Feature Scaling	Missing genre imputation
4	Time Series	Stock Price Forecasting	ARIMA, GARCH, ADF Test, ACF/PACF	Volatility modeling
5	Regression	Wine Quality Prediction	Linear Regression, Random Forest, Feature Importance	Model comparison
6	Regularization	Ridge & Lasso Regression	RidgeCV, LassoCV, Feature Selection	Baseball salary prediction
7	Classification	SVM Social Purchase	SVM, Grid Search, Decision Boundaries	Visualization of SVM regions
8	Vector Databases	Semantic Search with Pinecone	Sentence Transformers, Pinecone, Cosine Similarity	Weighted semantic search
9	Large Language Models	LLM Text Generation & Chatbot	HuggingFace Transformers, GPT-2, Qwen2.5	Creative generation, Q&A
10	R Programming	Housing Price Regression	Linear Regression, T-Tests, VIF Analysis	Statistical rigor in R
11	LangChain (Stub)	QA Chatbot with LangChain	(Design doc -- implementation in progress)	Coming soon

Technology Stack

Languages & Environments

Technology	Usage
Python 3.10+	Primary language for all ML/DL projects
R 4.x	Statistical analysis (housing regression)
Jupyter	Interactive notebooks for all projects
Bash	Automation and setup scripts

Machine Learning & Statistics

Library	Usage
scikit-learn	Classification, regression, PCA, SVM, model evaluation
statsmodels	ARIMA, ADF tests, statistical modeling
pmdarima	Auto-ARIMA for time series
arch	GARCH volatility modeling

Deep Learning & LLMs

Library	Usage
TensorFlow / Keras	Neural network for churn prediction
HuggingFace Transformers	GPT-2 generation, Qwen2.5 chatbot
Sentence Transformers	Embeddings for vector search

Vector Databases

Technology	Usage
Pinecone	Vector index for semantic search

Data Processing & Visualization

Library	Usage
pandas	Data manipulation and analysis
numpy	Numerical computing
matplotlib	Static and interactive visualizations
seaborn	Statistical data visualization

DevOps & Quality

Tool	Usage
Git / GitHub	Version control, CI/CD, project management
GitHub Actions	Automated linting, testing, security checks
pre-commit	Code quality hooks (formatting, linting, secrets)
Ruff / Black	Python linting and formatting
EditorConfig	Cross-editor consistency

Key Highlights

Statistical Rigor

Hypothesis testing (Levene's test, t-tests) in R
Model diagnostics (VIF, residual analysis, Q-Q plots)
Confidence intervals and p-value interpretation

Scalable Search

Vector embeddings with Sentence Transformers
Pinecone vector database for similarity search
Weighted query strategies for relevance tuning

Production-Grade Code

Clean Code principles
Consistent naming conventions
Comprehensive documentation
CI/CD pipeline with automated quality checks

Reproducibility

requirements.txt with pinned versions
Clear installation instructions
Self-contained Jupyter notebooks with markdown explanations

Architecture

data-science-portfolio/
|
+-- projects/                          # Project source code
|   +-- natural-language-processing/   # NLP projects
|   +-- deep-learning/                 # Deep learning projects
|   +-- machine-learning/              # ML projects (regression, classification)
|   +-- time-series/                   # Time series analysis
|   +-- llm/                           # Large Language Model projects
|   +-- vector-databases/              # Vector search projects
|   +-- r-programming/                 # R statistical analysis
|
+-- docs/                              # Additional documentation
|   +-- architecture.md                # Architecture overview
|
+-- scripts/                           # Utility scripts
|   +-- setup.sh                       # Environment setup script
|
+-- .github/                           # GitHub configuration
|   +-- workflows/ci.yml               # CI/CD pipeline
|   +-- ISSUE_TEMPLATE/                # Issue templates
|   +-- PULL_REQUEST_TEMPLATE.md       # PR template
|
+-- .gitignore                         # Git ignore rules
+-- .editorconfig                      # Editor settings
+-- .pre-commit-config.yaml            # Pre-commit hooks
+-- pyproject.toml                     # Project metadata
+-- requirements.txt                   # Python dependencies
+-- Makefile                           # Automation commands
+-- README.md                          # This file
+-- LICENSE                            # MIT License
+-- CHANGELOG.md                       # Version history
+-- CONTRIBUTING.md                    # Contribution guide
+-- CODE_OF_CONDUCT.md                 # Community guidelines
+-- SECURITY.md                        # Security policy

Installation

Prerequisites

Python 3.10 or higher
R 4.x (for the housing regression project)
Git
(Optional) Pinecone API key for vector database projects

Setup

# 1. Clone the repository
git clone https://github.com/AlfredBIAM/data-science-portfolio.git
cd data-science-portfolio

# 2. Create and activate a virtual environment
python -m venv venv

# Windows
venv\Scripts\activate

# Linux/Mac
source venv/bin/activate

# 3. Install dependencies
pip install -r requirements.txt

# (Optional) Install all optional extras
pip install -e ".[dev,nlp,deep-learning,time-series,vector-db,llm]"

# 4. (Optional) Set up pre-commit hooks
pre-commit install

Configuration (Vector Database Projects)

Copy the environment template and add your Pinecone API key:

cp .env.example .env
# Edit .env and add: PINECONE_API_KEY=your-key-here

Usage

Running Notebooks

# Start Jupyter
jupyter notebook

# Or with Jupyter Lab (recommended)
jupyter lab

Navigate to any project under projects/ and run the notebook cells in order.

Running the R Script

Open projects/r-programming/housing-regression/Linear_Regression_Housing_Data.R in RStudio or run:

Rscript projects/r-programming/housing-regression/Linear_Regression_Housing_Data.R

Building the LangChain Project (Coming Soon)

cd projects/llm/intro-to-llms/langchain-chatbot
pip install -r requirements.txt
python chatbot.py

Project Structure

For a detailed view of each project, see the individual README files:

Directory	Description
projects/natural-language-processing/fake-news-detection/	NLP pipeline for fake news classification
projects/deep-learning/customer-churn-prediction/	TensorFlow neural network for churn
projects/machine-learning/music-genre-classification/	PCA-based genre classification
projects/machine-learning/wine-quality-prediction/	Regression models for wine quality
projects/machine-learning/ridge-lasso-regression/	Regularized regression techniques
projects/machine-learning/svm-social-purchase/	SVM classification with grid search
projects/time-series/stock-price-forecasting/	ARIMA and GARCH forecasting
projects/llm/intro-to-llms/	HuggingFace text generation and chatbot
projects/vector-databases/semantic-search/	Pinecone vector search
projects/r-programming/housing-regression/	R statistical analysis

Development

Code Quality

# Lint all Python code
make lint

# Format all Python code
make format

# Run all checks (lint + format check)
pre-commit run --all-files

Conventional Commits

This project follows Conventional Commits:

feat: add new feature
fix: resolve bug in X
docs: update README
refactor: restructure Y module
test: add tests for Z
chore: update dependencies

Git Branch Strategy

main        -- Production-ready code
  develop   -- Integration branch
    feat/*  -- New features
    fix/*   -- Bug fixes
    docs/*  -- Documentation

Roadmap

Complete the LangChain QA chatbot implementation
Convert notebooks to Python scripts for production deployment
Add unit tests with pytest for data processing functions
Integrate MLflow for experiment tracking
Add Docker support for containerized execution
Deploy selected models as REST APIs with FastAPI
Add more comprehensive visualizations and dashboards
Implement model monitoring and drift detection

Contact

Kwami Alfred Jordal BIAM -- Data Scientist & AI Engineer

This portfolio represents a selection of academic and self-directed projects demonstrating competency in data science, machine learning, and artificial intelligence.

Name		Name	Last commit message	Last commit date
Latest commit History 2 Commits
.github		.github
docs		docs
projects		projects
scripts		scripts
.editorconfig		.editorconfig
.env.example		.env.example
.gitignore		.gitignore
.pre-commit-config.yaml		.pre-commit-config.yaml
CHANGELOG.md		CHANGELOG.md
CODE_OF_CONDUCT.md		CODE_OF_CONDUCT.md
CONTRIBUTING.md		CONTRIBUTING.md
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
SECURITY.md		SECURITY.md
pyproject.toml		pyproject.toml
requirements.txt		requirements.txt

Folders and files

Latest commit

History

Repository files navigation