Projects
Improving Malignancy classification in Lung nodule image through Curriculum Learning
Context encoder for artifact removal
Explainable AI - GradCAM Visualizations
Refined CNN based on Deep Feature Factorization
Principal Component Analysis, EDA on E-Commerce website
Energy Consumption-XGBoost, Decision tree
WHO Data Visualization: EDA on Human mortality
Supermarket Data Visualization: Time series Data analysis
Project
tour
Improving Malignancy classification in Lung nodule images
Energy Consumption-XGBoost, Decision tree
Created a classification model and predicted energy consumptions of buildings (EUI-High/Low) using XGBoost and Decision tree with confidence interval of 84% ±2%. Used XGBoost to further improve testing accuracy of the model.
Principal Component Analysis, EDA on E-Commerce website
Performed Exploratory data analysis ,dimensionality reduction on E-commerce website to find the factors influencing User's recommendations on shopping website using PCA . Utilized geo pandas to plot the users location to categorize them by various attributes.
Crop Disease Detection Using Convolutional Neural Network Algorithm
Created a mobile application to identify plant diseases using Android studio, TensorFlow, Keras frameworks. Leveraged TensorFlow, Keras frameworks to train a convolutional neural network model using adjusted mobilenetv2 (Google created opensource) with a dataset of 85,000 leaf images.Deployed the mobile application using Android studio as a disease detector app to give recommendations to plant diseases (preloaded) with 91percent accuracy
 |
|---|
POWER BI PROJECTS
Showcase of Work
WHO DATA VISUALIZATION
Data Source: Click Here
Performed exploratory analysis on WHO data: leading cause of disability-adjusted life year (DALY) & Death dataset, and visualized the metrics to be tracked. Created necessary slicers to slice and dice death due to diseases data across the globe by country, disease type, gender and year.
Supermarket Data Visualization
Data source: Click here
Performed predictive Analytics on Supermarket data to forecast total Sales data . Utilized Analytics Pane, Visualized Time series data by calculating Rolling Total and used ARIMA model to predict the sales by the end of next 3 months