🧠 Artificial Intelligence Labs

Welcome to the Artificial Intelligence Labs repository! This repository serves as a practical, hands-on portfolio showcasing end-to-end Machine Learning (ML) and Deep Learning (DL) pipelines.

The primary focus here is to move beyond isolated Jupyter Notebooks. These labs demonstrate the complete AI lifecycle: from raw data ingestion and Exploratory Data Analysis (EDA), through model training and hyperparameter tuning, all the way to model serialization and REST API deployment.

---

📑 Table of Contents

1. Core Methodology
2. Project Breakdown
3. Tech Stack & Tools
4. Repository Structure
5. Getting Started
6. API Deployment
7. Author

---

🏗️ Core Methodology

Each project in this repository generally adheres to the following workflow:

1. Data Preprocessing: Handling missing values, outlier detection, and data normalization/standardization.
2. Feature Engineering: Encoding categorical variables and selecting optimal features.
3. Model Development: Training baseline models and scaling up to complex algorithms (Ensemble methods, CNNs).
4. Evaluation: Utilizing accuracy, precision, recall, F1-score, and confusion matrices to validate model performance.
5. Deployment: Wrapping the optimized, serialized model (Pickle/Joblib or H5) into a FastAPI backend to serve real-time predictions via HTTP endpoints.

---

🔬 Project Breakdown

🧠 Deep Learning & Computer Vision

• 🫁 Chest X-Ray Classifier (COVID-19 & Pneumonia): A Convolutional Neural Network (CNN) trained on medical imaging to detect and classify lung infections.
• 🦠 Malaria Cell Classification System: An automated diagnostic tool using deep learning to identify parasitized cells in blood smear images.
• 🖼️ CIFAR-10 Object Recognition: A multi-class image classification model predicting objects across 10 distinct categories.
• 🔢 MNIST Digit Classifier: A foundational computer vision project recognizing handwritten digits with high accuracy.

📊 Machine Learning (Classification & Regression)

• 🏥 Diabetes Diagnostic Tool: A predictive classification pipeline to determine the likelihood of diabetes based on medical predictor variables.
• 🏠 House Price Predictor & Real Estate Estimator: End-to-end regression pipelines utilizing feature scaling to predict property values.
• 💰 Loan Approval Prediction: A financial risk assessment model classifying whether a loan should be approved based on applicant history.
• 🌸 Iris Species Classifier: A classic machine learning implementation establishing a solid baseline for multi-class prediction.
• 📱 SMS Spam Filter: Natural Language Processing (NLP) techniques applied to classify text messages as spam or legitimate.
• 💼 Salary Predictor: A regression model analyzing income trends based on experience and feature data.

---

🛠️ Tech Stack & Tools

• Programming: Python
• Data Manipulation & EDA: NumPy, Pandas, Matplotlib, Seaborn
• Machine Learning: Scikit-learn, XGBoost
• Deep Learning: TensorFlow, Keras
• Web Framework / API: FastAPI, Uvicorn
• Environment & Version Control: Jupyter Notebook, Git
• Infrastructure / DevOps: Docker, Windows Subsystem for Linux (WSL)

---

📂 Repository Structure

Artificial-Intelligence-Labs/
│
├── Chest X-Ray Classifier COVID-19 Pneumonia & Normal Detection/
├── CIFAR-10 Object Recognition/
├── Diabetes Diagnostic Tool/
├── House Price Predictor/
├── Iris Species Classifier/
├── Loan Approval Prediction/
├── Malaria Cell Classification System/
├── MNIST Digit Classifier/
├── Real State Price Estimator/
├── Salary Predictor/
├── SMS Spam Filter/
│
├── main.py                   # FastAPI application entry point (if centralized)
├── requirements.txt          # Python dependencies
└── README.md                 # Project documentation

---

🚀 Getting Started

Prerequisites

• Python 3.8 or higher installed.
• Git installed on your local machine.

Installation

1. Clone the repository:

   git clone [https://github.com/RealHaroon/Artificial-Intelligence-Labs.git](https://github.com/RealHaroon/Artificial-Intelligence-Labs.git)
   cd Artificial-Intelligence-Labs

2. Create and activate a virtual environment:

   python -m venv venv
   
   # On Windows:
   venv\Scripts\activate
   # On Linux/WSL/macOS:
   source venv/bin/activate 

3. Install dependencies:

   pip install -r requirements.txt

4. Explore the Notebooks: Launch Jupyter to view the training processes:

   jupyter notebook

---

🌐 API Deployment

Many of these models are designed to be production-ready. To spin up the FastAPI server and test the model endpoints locally:

1. Ensure your virtual environment is active.
2. Run the Uvicorn server:

   uvicorn main:app --reload

3. Open your browser and navigate to http://127.0.0.1:8000.
4. Interactive Docs: FastAPI automatically generates interactive API documentation. Visit http://127.0.0.1:8000/docs to test the endpoints directly from your browser.

---

👨‍💻 Author

Muhammad Haroon Abbas Khan BS Computer Science, University of Sindh, Laar Campus Badin

I am a Data Scientist and AI/ML developer focused on bridging the gap between raw data and deployed applications. I specialize in building end-to-end data pipelines, designing deep learning architectures, and serving intelligent web applications using robust relational databases (like Oracle and PostgreSQL) and modern API frameworks.

Certifications:

• 🏆 Oracle Cloud Infrastructure (OCI) AI Foundations Certified
• 🏆 Oracle Cloud Infrastructure (OCI) Generative AI Professional Certified

Connect with me:

• GitHub: @RealHaroon