Summer is the perfect time to dive into research, and you don’t need a lab or prior experience. All what you need is willingness to learn and background in programming and machine learning.
If you’re interested in data science or machine learning, here are three research projects you can start this week, inspired by real student work in our program.
1. What Factors Really Lead to NBA Wins?
Focus: Sports Analytics + Predictive Modeling
Skills you’ll learn: Feature selection, classification, model evaluation
A student once asked: “What actually causes a basketball team to win?” Is it shooting percentage? Rebounds? Assists? Defense? Instead of guessing, they analyzed thousands of NBA games using machine learning to find patterns that lead to a win.
Your challenge: Use NBA game data to build a model that predicts whether a team will win based on features like 3-point percentage, turnovers, rebounds, and assists. Then identify which factors matter most, and how they differ between winning and losing teams.
Dataset: Kaggle: NBA Games Stats Dataset
2. Can You Use MRI Scans to Help Detect Brain Tumors?
Focus: Medical Imaging + Computer Vision
Skills you’ll learn: Image classification, convolutional neural networks (CNNs), biomedical AI ethics
Medical research may sound out of reach, but you can actually use public MRI scan datasets to train models that detect signs of brain tumors. One of our students used AI to build models to classify and predict cancer patients.
Your challenge: Use a deep learning model to classify MRI images as showing a tumor or not. Then evaluate how accurate your model is, and what challenges arise when applying AI to medicine.
Dataset: Kaggle: Brain Tumor MRI Dataset
3. Can You Predict Whether a Product Will Be Popular on Amazon?
Focus: E-commerce + Text & Data Analysis
Skills you’ll learn: Classification, NLP, sentiment analysis, feature engineering
Thousands of products are launched every month on Amazon, some become hits, others flop. This is inspired from a student project that analyzed the features of products’ pages to gain insights on what makes products popular.
Your challenge: Analyze Amazon product data to build a model that predicts whether a product will be rated 4 stars or higher. Use product titles, categories, pricing, and review sentiment as your features.
Dataset: Kaggle: Amazon Product Review Dataset
Get Started Today!
These projects are beginner-friendly but grounded in real-world applications. They’re the kind of work that can lead to published papers, awards, and standout college applications, and they’re all doable from your laptop!
Do you know a high school student who wants to work on a research project? Forward this article to them!
Do you want to share your progress and get feedback from other students and mentors about your research? Join our skool community: https://www.skool.com/research-college-prep-5443