Nosa Omokaro Portfolio

Credit Card Fraud Detection

I developed a machine learning pipeline to detect fraudulent transactions in a highly imbalanced dataset. I implemented class balancing techniques, feature scaling, and optimized recall and precision to minimize false positives and false negatives. This project highlights my expertise in fraud detection, handling class imbalance, and building machine learning models, achieving over 90% accuracy and recall.

View Project

Customer Segmentation & Clustering Analysis Project

In this project, I Designed and implemented a clustering analysis project using Python to segment customers from the Mall Customers dataset based on annual income and spending habits. Leveraged tools like Pandas, Scikit-learn, and Matplotlib to preprocess data, apply KMeans clustering, and visualize results. Identified five distinct customer segments using feature scaling and the elbow method, providing actionable insights into customer behavior. The project demonstrated proficiency in data preprocessing, machine learning algorithms, and effective communication of findings through visualizations and exported datasets for further analysis.

View Project

Automated Job Evaluation System Using Machine Learning

I developed a machine learning-based system to automate job description evaluations within the Higher Education Role Analysis (HERA) framework. Implemented and compared various text classification models, including Logistic Regression, SVM, and Random Forest, achieving a 93.8% accuracy in role classification. Addressed class imbalance and ensured model interpretability, leading to a more consistent and fair HR process. Technologies Used: Python, Scikit-learn, NLTK, Random Forest, Logistic Regression

View Project

Data Cleaning in SQL

In this project, I modified Nashville Housing data in SQL Server to improve its analytical use

View Project

Data Exploration in SQL

In this project I explored global COVID 19 data using SQL Server with which I derived meaningful insights from the dataset

View Project

London Housing Price Prediction Model

In this project I evaluated machine learning models for predicting house prices in London, leveraging a comprehensive dataset (London Housing Prices). Four distinct models were employed: Linear Regression, Random Forest Regressor, Elastic Net, and a Neural Network. The models underwent thorough evaluation using various metrics, providing insights into their performance and suitability for the given task

View Project

Stroke Data Management System project

In the Comprehensive Stroke Data Management System project, I served as a Data Engineer, Data Analyst, and Database Developer. I developed and optimized SQL and NoSQL databases, designed data pipelines, and implemented a scalable infrastructure using Azure Databricks. My work included conducting data analysis to extract insights on stroke demographics and trends, and creating interactive visualizations in Power BI to support data-driven decision-making.

View Project

Addidas U.S Sales Project

In this project I explored Addidas U.S Sales Data using Sql, transformed and cleaned the data using DAX and Excel, and created a Power Bi dashboard to identify the top performing products. I also provided insights on maximizing the utilization of the data.