Vishwas Pai

Experienced Data Analyst with 1.3 years of expertise in analyzing data and deriving actionable insights. Demonstrated strong interpersonal, leadership, and management skills, with a proven ability to communicate complex technical concepts effectively. Eager to learn and apply new technologies and tools to drive data-driven decision-making and enhance operational efficiency.

Methane emission estimation from livestock model - based on livestock characteristics and feed nutritional profiles regression model.

• extracting data from MongoDB, EDA, data preprocessing, IQR method for outliers, encoding categorical features, checking Pearson's correlation, VIF factor for multicollinearity, Model selection, joblib for model saving.
• ExtraTreesClassifier final accuracy score - 95.0% and AUC score - 0.99

Insurance claims - Fraud detection model, Building a classification model to predict if an auto insurance claim is fraudulent or not.

• EDA, data visualization, handling outliers using Z-Score method, removed Multicollinearity, treated data imbalance using SMOTE, built multiple classification models, performed Hyperparameter tuning using GridSearchCV, created confusion matrix and AUC ROC curve, saved best model for prediction.,
• Linear Multiple Regression R2 score - 96.10% and CV score - 95.97%

Used Car price prediction model - Building a Regression model to predict the prices of used cars.

• Solution: EDA, data preprocessing, feature engineering, imputing null or missing values, Z-Score, removing skewness, encoding categorical features, standard scaler, VIF, feature Importance graph, Model selection, GCV on best model.
• Extreme Gradient Boosting R2 score - 97.55% and CV score - 94.92%