MAYANK VISHAL

Qualified Data Science Intern with solid foundation in statistical analysis and machine learning. Contributed to predictive modeling projects and data visualization tasks in previous internship, demonstrating practical applications of theoretical knowledge. Proficient in Python and R, with keen ability to interpret complex data sets and collaborate effectively within team environments.

• SQL for Data Science, Ucdavis (Coursera), 2024
• Creating Dashboards and Storytelling with Tableau, Ucdavis (Coursera), 2024
• Excel Skills for Business: Advanced, Macquarie University (Coursera), 2023
• CBAP Certified, Koed, 2024

• Participated in the RSNA 2024 competition to develop models for diagnosing degenerative spine conditions using lumbar spine MR images, aiming to simulate radiologist performance in classification.
• Preprocessed MRI data applied image augmentation, and fine-tuned CNN models (e.g., ResNet, EfficientNet) for multi-label classification. Used cross-validation and model ensembling to enhance performance.
• Achieved top 10% placement, showcasing the effectiveness of AI models in assisting radiologists with accurate and efficient diagnosis of degenerative spine conditions.

Image Recognition Using OpenCV and Machine Learning:

• · High Classification Accuracy: Achieved a classification accuracy of 85%-95%, evaluated using performance metrics like confusion matrix, precision, recall, F1-score, and ROC curves for comprehensive assessment.
• · Effective Preprocessing: Preprocessed input images with OpenCV techniques including resizing, grayscale conversion, normalization, and edge detection, ensuring robust feature extraction and reducing noise for model inputs.
• · Feature Engineering and Model Building: Utilized extracted features to train machine learning models, including SVM, Random Forest, and CNN architectures, optimizing hyperparameters to improve model performance.

Life Expectancy Prediction Model Objectives:

• · Indicator Analysis: Analyzed health, economic, and environmental indicators (e.g., GDP, healthcare access, pollution levels) to identify key factors influencing life expectancy.
• · Data Preprocessing: Addressed missing data by employing imputation techniques such as mean substitution and regression-based approaches, achieving a 95% reduction in missing values for cleaner datasets.
• · Model Development: Created and evaluated machine learning models like Linear Regression, Random Forest, and Gradient Boosting, achieving a Mean Squared Error (MSE) of 3.25 and an R² score of 0.88 (88% accuracy).

Stock Price Prediction using ML Model

• · Data-Driven Prediction: Built a Linear Regression Model (LRM) in Jupyter Notebook to predict Reliance stock prices, focusing on robust preprocessing, including feature engineering, normalization, and outlier handling.
• · Model Training & Validation: Split data into training/testing sets and applied cross-validation for model robustness, ensuring accurate predictions.
• · Performance Metrics: Achieved Mean Squared Error (MSE) of 2.84, Mean Absolute Error (MAE) of 1.65, and a prediction accuracy of 90%.

Managed Printing Services Implementation

• · Efficiency Audit: Conducted a thorough analysis of printer usage, identifying inefficiencies and proposing strategies to optimize cost and resource use.
• · Solution Development: Researched and evaluated MPS frameworks and vendor solutions for scalability, operational efficiency, and environmental impact.
• · Business Case Preparation: Created a detailed cost-benefit analysis with ROI projections to support management decision-making.
• Sustainable Impact: Achieved a 20% reduction in printing costs, minimized paper waste, streamlined workflows, and supported sustainability goals through better resource management.