Engaged in a collaborative advanced machine learning research project focused on analyzing genomic data for uterine cancer, utilizing the comprehensive TCGA-UCEC dataset from the National Cancer Institute's Genomic Data Commons (GDC) Portal to identify potential diagnostic biomarkers.Designed and developed a Convolutional Neural Network (CNN) to investigate the impact of genetic mutations (including LATS1, LATS2, MST1, YAP1, and TAZ) on the detection and diagnosis of uterine cancer. Employed cutting-edge deep learning techniques to enhance model accuracy and predictive performance.Managed and processed medical imaging data, specifically DICOM images, which involved systematic labelling and preparation of data sets to ensure the integrity and usability of data for effective model training and validation.Achieved exceptional results, with the CNN model demonstrating a notable test accuracy of 95.77% in detecting uterine cancer based on the LATS1 gene mutation, significantly advancing the potential for early diagnostic applications and personalized medicine approaches in oncology.

Managed and organized the collection of a comprehensive dataset including approximately 10,000 customers, capturing a wide array of characteristics and behavioural data spanning the previous year, to support targeted analytics initiatives.Executed thorough data preprocessing tasks, including cleaning, imputing missing values, and transforming categorical variables into numerical formats, to ensure high-quality data for analysis.Performed extensive Exploratory Data Analysis (EDA) to uncover underlying patterns and relationships between various customer attributes and their churn rates, gaining insights that guided further analytical efforts.Tested and evaluated multiple advanced machine learning algorithms, such as Logistic Regression, Support Vector Machines, and Random Forests, utilizing robust Cross-Validation techniques to determine the optimal model based on performance metrics.Developed and refined a Gradient Boosting Classifier model, which achieved an impressive accuracy of 85% and an AUC-ROC score of 0.89, demonstrating significant predictive power and reliability in identifying potential customer churn.

Supported the analytics team in examining complex business issues by performing detailed data analyses using Powerview and Worldpanel Online tools, which involved extracting actionable insights from diverse datasets tailored to client-specific needs.Developed and presented visual data representations using Tableau, enhancing the interpretability of analytical results, which facilitated strategic decision-making. Contributed to the preparation and refinement of client reports, ensuring clear communication of findings and recommendations.Engineered data solutions to streamline operations, including designing and coding APIs in Python to enhance data accessibility and integration. Crafted a user-friendly interface using React and JavaScript, significantly improving the interaction experience for end-users.

Rapidly mastered Power BI tools from scratch, facilitating comprehensive business data analysis by integrating multiple data sources such as SQL Server and Excel within two months. Executed data integration and cleansing processes to ensure accuracy and usability.Leveraged Power BI to develop an interactive supply chain inventory status analysis report, utilizing Power Query for data manipulation and DAX for advanced calculations, thereby enhancing data visualization and providing insightful operational intelligence to support strategic decision-making.