Greetings! I'm Sneha Karki, a Data Science grad student at Michigan Tech, blending creativity with technical prowess. From predicting energy consumption to diving into TV analytics, I love solving challenges. Beyond data, I'm a storyteller and problem solver. Outside of studies, I enjoy writing, watching movies, and meeting new people to hear their stories. Let's connect and explore the intersection of data, innovation, and success! 🚀

Current Favorite Quote:

"In three words I can sum up everything I’ve learned about life: It goes on."
-Robert Frost

My Skills

Programming Skills

Python

Data analysis Skills

Data Cleaning

Data Visualization

Statistics

Software and Database Skills

SQL

HTML

CSS

Projects

Fake News Detection using Voting Technique

This project tackles fake news using advanced machine learning with a unique voting method, outperforming individual classifiers. Results across diverse datasets showcase its effectiveness in identifying fake news, detailed in the GitHub repository.

Analysis of Streaming Shows

This project involves analyzing shows and movies on popular streaming platforms using Python libraries for both data visualization and analysis. From the dataset, valuable insights were obtained regarding platform and show popularity.

Appliances Energy Consumption

This project predicts household energy consumption using diverse regression models and a Kaggle dataset. Focused on key areas, it includes preprocessing, exploratory data analysis, and model application. Top models, Principal Component Regression and Logistic Regression, achieved notable performance metrics.

Chronic Kidney Disease prediction

This project tackles Chronic Kidney Disease using machine learning, predicting outcomes and identifying key risk factors like diabetes and hypertension. With accuracy rates up to 90%, the Random Forest model outperformed others. Findings contribute to healthcare knowledge, aiding informed decision-making, and suggesting interventions for policymakers.

Predicting Mental Health Diagnosis in Academia

This project uses machine learning to predict mental health diagnoses among university students, achieving 82% accuracy. The model identifies 35 key survey questions, offering an efficient tool for early identification and support in higher education.