Data wrangling, solution design, model building and communicating results.

• Developed an app to minimize screening time for Covid19 using chest radiographs.• Built a sequential model using Keras tensorflow library to distinguish Covid19 from pneumonia and normal radiograph using Softmax activation function.• Performed data augmentation using techniques like: rotation, translation, zooming, cropping, flipping and changing brightness.• Implemented Alex-net architect for CNN model and acquired Specificity of 87% with limited amount of Covid19 chest X-ray… Show more • Developed an app to minimize screening time for Covid19 using chest radiographs.• Built a sequential model using Keras tensorflow library to distinguish Covid19 from pneumonia and normal radiograph using Softmax activation function.• Performed data augmentation using techniques like: rotation, translation, zooming, cropping, flipping and changing brightness.• Implemented Alex-net architect for CNN model and acquired Specificity of 87% with limited amount of Covid19 chest X-ray data. Show less

• Developed a binary classification model using XGBoost to predict whether a person would develop coronary heart disease in 10 years.• Utilized SMOTE with random under-sampling to generate a balanced training dataset.• Performed data preprocessing - Outlier Treatment, Missing Value treatment and Engineered features like : Hypertension class, Diabetes severity and Smoking factor.• Early prediction of coronary heart disease can help to prevent the advancement of the disease and avoid… Show more • Developed a binary classification model using XGBoost to predict whether a person would develop coronary heart disease in 10 years.• Utilized SMOTE with random under-sampling to generate a balanced training dataset.• Performed data preprocessing - Outlier Treatment, Missing Value treatment and Engineered features like : Hypertension class, Diabetes severity and Smoking factor.• Early prediction of coronary heart disease can help to prevent the advancement of the disease and avoid more invasive treatment procedures - Achieved a recall rate of 92%. Show less

• Developed an App to detect the identity of a student during online class and build automated attendance system. Key features include an alarm to alert for drowsiness.• Extracted face region using OpenCV with Haar Cascade classifier and resized images with augmentations like: rotation, translation, flipping and changing brightness.• Built a Siamese neural network to recognize a person's face in one shot learning using 136 face encodings, with a time-based algorithm separated blink from… Show more • Developed an App to detect the identity of a student during online class and build automated attendance system. Key features include an alarm to alert for drowsiness.• Extracted face region using OpenCV with Haar Cascade classifier and resized images with augmentations like: rotation, translation, flipping and changing brightness.• Built a Siamese neural network to recognize a person's face in one shot learning using 136 face encodings, with a time-based algorithm separated blink from drowsiness.• Deployed the web application built using Streamlit on cloud - AWS EC2 instance. Show less

Consulting, Treatment plans and managing patients experience through their life cycle.

1.3k+ Answers with an overall CF Score of 96%.DFS, BFS, Binary Search Tree, Linked List, Object-Oriented Programming, Inheritanceand Polymorphism, Dynamic Programming, Concurrency, etc.

HOD Dental Department, General, and Oral Health Check-Up Camp.

Consultation, Treatment planning and Providing Dental treatment.