Ekta Bhadouria

• Implemented media mix modelling to quantify contribution of various marketing channels on overall sales
• Developed comprehensive framework for data collection, preprocessing, model training and for result using azure data storage
• Utilized prophet model for trend seasonality, lightweight MMM model for MMM project
• Helped in automating hyperparameter tuning pipeline which is developed on kubeflow and automating result
• Developed anomaly detection model using bidirectional LSTM

A/B testing Project: Performed the A/B testing on the various control test pair and conclude the result

• Forecasting models for inventory forecast
• Implemented various advance models like ARIMA, prophet, tbats, SARIMAX etc., along with different visualization of data required by business
• Linear Optimization model for better managing inventory for one of the leading brewery company
• Proposed solution for cost function and understand constraints to define problem statement for optimization.

• Developed visualization tool for data analysis
• Developed Machine learning (Decision Tree, Random Forest, KNN) and Deep learning-model (ANN) algorithm (pipeline to find cognitive load on the driver based on parameter provided under Jaguar Land Rover research project
• Deployed project using flask framework
• Created a pipeline to deploy project on AWS cloud.

• Application developed with the data from Affectiva SDK to find the emotion and created the prototype application for the demonstration purpose Research, Data analysis and model development to find the emotion of driver
• Developed closed loop application with the integration of YouTube API
• Performed complete life cycle of Data Science Project on various open-source dataset Management Internal project on classification, regression, logistic algorithm
• Actively worked as a self-learned developer and gave training to interns.

• MATLAB based model for jerk minimization
• Simulink model validation.