Jaswanth Mada
Jaswanth Mada
Master's in Computer Science, Purdue University | Machine Learning & AI • Software Engineering
- AI/ML Engineer and Full-Stack Developer with a passion for solving real-world problems.
- Strong competitive programming with excellent problem-solving abilities and analytical thinking.
- Experience building machine learning models using PyTorch and TensorFlow
- Skilled in Python, JavaScript, React, and Node.js.
- Built scalable web applications with MongoDB and AWS.
India | United States
Projects
Professional Experience
Skills
Education
Competitive Coding
Certifications
DS/ML Projects
Full Stack Projects
Frontend Projects
Mobile Price Prediction
Problem
Smartphone pricing varies widely due to brand premiums, regional differences, and rapid depreciation, making it difficult for consumers and businesses to make informed pricing decisions.
Credit Card Fraud Detection
Problem
Credit card fraud costs $24+ billion annually, but fraudulent transactions make up only 0.17% of all transactions. The challenge is detecting rare fraud patterns while avoiding false positives that disrupt legitimate customers.
Brain Tumor Classification with Fine-Tuned VGG16
Problem
Developed a deep learning model to accurately classify brain MRI scans, addressing the challenge of limited medical imaging data through an effective transfer learning pipeline.
Credit Card Fraud Detection
Problem
Developed a machine learning pipeline to detect fraudulent credit card transactions, addressing extreme class imbalance in a dataset of 284K transactions with only a 0.17% fraud rate[cite: 29].
Cat vs. Dog Classifier
Problem
Developed a deep learning model to accurately distinguish between images of cats and dogs, a foundational computer vision binary classification task.
House Price Prediction
Problem
Developed a machine learning model to accurately predict residential house prices using a variety of features, demonstrating a mastery of the end-to-end regression modeling pipeline.
Email Spam Classifier
Problem
Developed an NLP-based machine learning model to automatically classify emails as 'spam' or 'ham' (not spam), demonstrating proficiency in text data processing and supervised learning.
Recommendation Systems
Problem
Developed and implemented a recommendation system to provide personalized item suggestions to users, enhancing user experience and engagement by solving the cold-start problem and data sparsity challenges.