Priyanka Kondi

Data Cleaning and Organization: Leveraged Excel to clean and structure large datasets, ensuring data integrity by correcting errors and excluding metro areas with fewer than 50,000 workers.

Data Analysis and Visualization: Used Excel to perform in-depth analysis, calculating workforce percentages, ranking job concentrations, and creating visual representations (charts/graphs) to illustrate workforce trends and distributions.

STEM Categorization: Classified occupations into STEM and non-STEM categories using SOC codes, enabling focused analysis of Huntsville’s job market strengths and areas for improvement.
Development of Reporting Tool: Created a user-friendly reporting tool to provide stakeholders with accessible insights, supporting strategic workforce development decisions.

Tools Utilized: Excel (for data cleaning, analysis, and visualization), Reporting Tools (for streamlined presentation of findings), SOC Codes (for occupational classification).

Data Collection and Cleaning: Sourced and cleaned transaction data, addressing missing values and outliers to ensure data quality using Python.

Exploratory Data Analysis (EDA): Conducted detailed EDA to uncover patterns and correlations in the data, leveraging Matplotlib and Seaborn for visualization.

Feature Engineering: Engineered new features, including time-based attributes and categorical encoding, to enhance model performance, using Pandas and NumPy.

Model Development: Built and optimized multiple machine learning models, such as Logistic Regression, Random Forest, and XGBoost, using scikit-learn.

Model Evaluation and Selection: Evaluated models with metrics like Accuracy, Precision, and ROC-AUC, selecting the most effective model using GridSearchCV from scikit-learn.

Implementation and Testing: Deployed the selected model in a real-time fraud detection system and tested it extensively for reliability and accuracy, using Python.