This is a private internal preview of DISCOVERY 2.0. Do not share externally.

Data Science MicroProjects

MicroProjects are detailed but small projects designed to do real Data Science in under an hour using real-world tools including GitHub, Python, Visual Studio Code, and more! (So far, we've awarded thousands of MicroProject cards -- we hope you'll complete the whole set! :))

#1: Trends in High School GPAs

Explore the high school GPAs of the incoming freshman classes of Big Ten schools while using row section with DataFrames and creating line charts.


Dataset: Common Data Set, CSV

#2: NWS Hourly Weather Forecast

Find the hourly weather forecast for any location in the United States while learning to use a JSON web API to fetch data into a DataFrame for analysis, and to create a scatter plot of your weather forecast.


Dataset: National Weather Service (NWS), JSON API

#3: Highest Mountains in the World

Transform data found in Wikipedia articles into DataFrames, combine multiple DataFrames together using concatenation, a build a bar chart from a DataFrame.


Dataset: Wikipedia URL, HTML tables

#4: United States Congress

Learn the international standard representation of dates and times (ISO 8601) by exploring the birthdays of every member of Congress, creating both a scatter plot and box plot along the way!


Dataset: @unitedstates GitHub, CSV

#5: Bechdel Test

Learn to create a DataFrame pivot table from a collection of over 10,000 films with Bechdel Test ratings, creating a stacked area chart to see how the representation of women in a film has changed over time.


Dataset: BechdelTest.com, JSON API

#6: Infinite Money in Roulette (Martingale Betting System)

Motivated by the claims of an "infinity money glitch", create a simulation of the game of Roulette and play out 1,000,000 games to find if it really works by visualizing the results using several line charts!


Dataset: Your own simulation in Python

#7: Creating an Image Mosaic in Python

Build you own unique image mosaic by representing image pixel data (x, y, r, g, b) as a DataFrame and find the best possible match for every region of an image.


Dataset: Your Own Image Files

#8: Image Steganography with DataFrames

Represent image pixel data in a DataFrame and follow a small treasure hunt by finding hidden images in data by decoding the hidden data we've left for you!


Dataset: Image Files w/ Hidden Data

#9: FIFA Matches since 1872

Learn to summarize datasets with panda's powerful groupby function to analyze over 45,000 FIFA matches, creating both a stacked bar chart and mutli-line line graphs.


Dataset: Mart Jürisoo (@martj42) Repository of FIFA Matches

#10: Chess and The Elo System

Explore over a hundred thousand real chess matches played on lichess.org to understand the Elo system used to rank players, and understand how to adapt simulation code to deterministic events.


Dataset: lichess.org open database, CSV

MicroProject #11: Learning Handwritten Digits with AI

Using the most famous dataset in machine learning, apply both K-Means Clustering and K-Nearest Neighbor to find the best algorithm to train an AI model to learn to recognize handwritten digits.


Dataset: Images of Handwritten Digits (MNIST)

#12: Building a Scene Recognition Model from Video Frames

Create your very own AI classifier using images captured from a Data Science DISCOVERY office hours video!


Dataset: Video Frames from DISCOVERY

#13: Choropleth Maps

Geographical data visualizations are some of the most impactful forms of data visualization, learn to use example code from the folium project to create a custom choropleth maps from our own CSV data!


Dataset: UIUC Enrollment by Home State, CSV

#14: AI vs Human: Response Analysis

Compare human and AI responses to the same questions and find if there are statistically significant differences between a human and AI response using Natural Language Processing (NLP) techniques.


Dataset: Hugging Face HC3 Dataset

MicroProjects from the Vault

Nerd out with our previous MicroProjects, using older datasets or not meeting our current standards, to explore more datasets and data science skills!

Vault #1: World University Rankings

From the DISCOVERY Vault: Practice basic data science with pandas while explore the top universities in the world with the dataset from Times Higher Education (THE).


Dataset: Times Higher Education Rankings, CSV

Vault #2: Illini Football Since 1892

From the DISCOVERY Vault: Nerd out with a live dataset of every Illini football game ever played, dating all the way back to 1892, and discover trends in over 130 years of data!


Dataset: Illini Football Dataset, CSV

Vault #3: Exploring COVID-19 Data from GitHub

From the DISCOVERY Vault: Explore data directly from GitHub that was updated daily by Johns Hopkins University's Center for Systems Science and Engineering (CSSE) every day as they monitored the spread of COVID-19.


Dataset: Johns Hopkins University COVID-19 Data, CSV

Vault #4: Simulation for Ten Heads in a Row

From the DISCOVERY Vault: Explore working with time series data by finding trends of sequential patterns of data through a simple simulation of flipping a fair coin! Can you get ten heads in a row?


Dataset: Python Simulation with DataFrames

Vault #9: United Nations (UNHCR) Refugee Data

Work with a web-based API from the United Nations.


Dataset: United Nations (UNHCR) Web API