We are Data Science at Georgia Tech, Georgia Tech's largest community of student data scientists. Our vision is to foster passion in data science and create a positive impact on the community. This passion inspired Hacklytics, our data-science themed hackathon. We encourage both experienced students and beginners to come together and build something new and impactful, while having fun!
Hacklytics is Georgia Tech’s 24 hour datathon brought to you by Data Science at Georgia Tech! Our
theme for this year is "Change the World" with a focus on sustainability and growth. We hope to see
fun, impactful and insightful projects that seek to bring change to society.
A datathon is a specific type of hackathon that focuses on data science. While this may make them
seem constricted, we choose to approach them with the same idea as a normal hackathon, but applied
to data science: whatever data you want, whatever language you want, whichever APIs you want,
whatever ML/DL algorithm to analyze it you want. You can create visualizations, make a model, come
up with insights and anything you find impactful!
Hacklytics will be held from February 22nd-23rd 2020 in the Klaus Advanced Computing Building,
(266 Ferst Dr NW, Atlanta, GA 30332).
You can fill out the form using the following link:
tiny.cc/hacklytics-apply.
Applications close February 1st!
We will evaluate your responses to our questions and interest in participating and making the
most of the datathon. (Don't worry - we won't be judging your writing skills!)
We will have mentors and volunteer forms up shortly. Keep an eye out for those!
Teams of up to 4 people are permitted. We will have a channel in Slack for other people who need
to find a team and we will allow for team-building time before the hacking begins.
Nothing at all! Hacklytics is free of cost for you to attend and we will be providing food and
hacking space.
Minors and non-students are not allowed to attend Hacklytics.
We have events and workshops as well as networking sessions planned out for you throughout the
day, so you have something to do while that model finishes training!
Not at all! We welcome students with varying skill levels, and we have workshops and sessions
planned for you to learn something. So whether you are a beginner or an experienced data scientist,
we will have something for you!
Any hacking gear you need (laptop, hardware, chargers, batteries, etc.), comfortable clothes,
toiletries (toothpaste, toothbrush, deodorant, etc.), a photo ID for registration, a government ID
to rent hardware, and most importantly, yourself!
Unfortunately this year we will not be able to provide travel reimbursements to students.
Contact us at hacklyticsgt2020@gmail.com!
All hackers must adhere to the MLH code of conduct.
Topic: General Economics
Website: https://data.worldbank.org/indicator/ny.gdp.pcap.cd
Description: Contains 9 different datasets that can be downloaded
as CSV, Excel, or XML files. Most of the data sets are different measures of GDP per
capita for a lot of countries (at least 200). It also contains a data set on inflation
and another on oil rents.
Topic: Socioeconomic factors
Website: https://opportunityinsights.org/data/
Description: This site probably contains any information you might
want about socioeconomic equality. There are around 50 datasets, A lot of them have to
do with economic mobility and how it relates to race, gender, income, and other factors.
In addition to economic mobility, datasets also cover geographical mobility and even the
role colleges play in economic success. All datasets can be downloaded as Excel
spreadsheets and there are READMEs associated with each one that talk about each field
in the dataset.
Topic: Homelessness
Website:
https://www.hudexchange.info/homelessness-assistance/ahar/#2019-reports
Description: This site has around 30 datasets if you click around
to find them. All these datasets are about homelessness. There are a few on here I think
are really useful. The 2018 CoC Point in Time estimates contains homelessness counts for
every county in the United States from 2007 to 2018. The 2019 section also contains data
on veteran homelessness including change over the years. All datasets can be downloaded
as Excel spreadsheets
Topic: Environment
Website: http://geodata.grid.unep.ch/
Description: This is a database of environmental datasets. You can
download all these datasets as either CSVs or Excel spreadsheets. The database is also
searchable making it easy to hone in on certain topics you might be interested in.
Unfortunately, it seems like the newest data on there is still a few years old, so it
might not be the best dataset for looking at new trends, but for looking at long term
trends from the past 50 years or so, this would be a good source.
Topic: Everything
Website: https://toolbox.google.com/datasetsearch
Description: Description: Use Google dataset search to find a
dataset on anything you want.
Topic: General Economics
Website: https://fred.stlouisfed.org/
Description: 627,000 searchable datasets about information from
unemployment to inflation to population statistics. You can not only download these
datasets as csv files or excel files, but also download graphs created by the data as an
image, PowerPoint, or PDF. This is probably one of the best sources you can get for
American economic data.
Topic: Socioeconomic factors
Website: https://data.humdata.org/
Description: This dataset focuses a lot of developing countries and
the problems they face. You can find information like access to electricity, access to
education. Datasets are usually available as CSV files and Excel files. Unfortunately,
you need to request access for some of the listings in this database. A lot of the data
also has accompanying reports in the form of PDFs and even some PowerBI’s that you can
use to make more sense of the data.
Topic: Environment
Website: https://www.climate.gov/maps-data/datasets
Description: There are around 50 different datasets and datamaps on
this site, all related to weather patterns and changes over time. Not all the data can
be downloaded as CSVs or Excel spreadsheets as there are some maps. In addition,
actually acquiring the dataset seems somewhat involved for a lot of the datasets listed.
There are how-to tabs that describe the process of getting a particular dataset, but
it’s a bit more than simply clicking a download link.
Topic: Everything
Website: https://catalog.data.gov/dataset
Description: This is a searchable database from the US Government
about everything. There are over 250,000 datasets in this database. Unfortunately, not
all of them are downloadable as CSV or Excel files, but there are enough datasets to be
useful here. You can search by tags as well. As a side note, I actually found this
database while looking for pollution datasets, so this might be pretty good if you are
looking for environment related data.
Topic: General NLP / Machine Learning
Website: https://quantumstat.com/dataset/dataset.html
Description: 200 datasets that all have descriptions on what they
are about. If you want to do an NLP project, this might be a good place to start. Most
datasets are available for download in either JSON, CSV formats.
Topic: Autonomous Vehicles
Website: https://level5.lyft.com/dataset/
Description: This is pretty different from the other datasets as it
contains annotated images taken from Lyft L5’s self-driving car fleet. There are over
55,000 human-labeled 3d annotated frames, data from cameras and lidars, and drivable
surface maps included.
Topic: NLP / Human-understanding AIs
Website:
https://storage.googleapis.com/dialog-data-corpus/TASKMASTER-1-2019/landing_page.html
Description: The dataset consists of 13,215 task-based dialogs,
including 5,507 spoken and 7,708 written dialogs created with two distinct procedures.
Each conversation falls into one of six domains: ordering pizza, creating auto repair
appointments, setting up ride service, ordering movie tickets, ordering coffee drinks
and making restaurant reservations. For more information, there is a readme on the
website. You’ll need to download the dataset using command line tools.
Topic: Everything
Website: https://github.com/TheUpshot
Description: These are open source data sets used by New York
Times' The Upshot.
Topic: Politics and Sports
Website: https://data.fivethirtyeight.com/
Description: These are open source data and code by
FiveThirtyEight.
Topic: Everything
Website: https://archive.ics.uci.edu/ml/index.php
Description: These 488 data sets provided by and maintaned by UC
Irvine and are intended for machine learning.
Topic: US Social Issues
Website: https://gss.norc.org/
Description: This data set includes survey data going back to 1972.
The survey covers the social opinions of Americans.
Topic: Everything
Website: https://data.worldbank.org/
Description: Search for data sets provided by the The World Bank.
Topic: Seattle City Data
Website: https://data.seattle.gov/
Description: The city of Seattle has open sourced its data on
education, finance, and various other topics.
Topic: Financial
Website: http://api.reimaginebanking.com/
Description: The is Capital One's hackathon API Nessie. This
contains data such as atm and bank branch locations.
| 2/22 - Saturday | 12:00 PM | Participant check-in | Klaus Atrium |
| 1:00 PM | Opening ceremony | CoC 016 | |
| 1:45 PM | Team formation | Klaus Atrium | |
| Sponsor expo | |||
| Work time begins | |||
| 2:00 PM | Workshop: Basics of Python | Klaus 1456 | |
| Late check-in closes | |||
| 3:00 PM | Workshop: Ideation and Prototyping | Klaus 1456 | |
| 4:00 PM | Workshop: Making Use of Data | Klaus 1456 | |
| Workshop: Dim Reduction and PCA | Klaus 2456 | ||
| 5:00 PM | Workshop: Unsupervised Learning | Klaus 1456 | |
| 5:30 PM | MLH: US Air Force CTF | Klaus 2456 | |
| 6:00 PM | Dinner! | Klaus 1116E | |
| 6:20 PM | The New York Times talk, featuring Andrew Marchese | Klaus 1456 | |
| 8:00 PM | Workshop: Predictive Analysis | Klaus 1456 | |
| 9:00 PM | Workshop: Deep Learning | Klaus 1456 | |
| 2/23 - Sunday | 1:00 AM | Midnight snack | Klaus 1116 W |
| 8:00 AM | Breakfast! | Klaus 1116 E | |
| 9:00 AM | Andrew Marchese Q&A | Klaus Atrium | |
| 10:00 AM | Workshop: Tableau | Klaus 1456 | |
| 12:00 PM | Lunch! | Klaus 1116 E | |
| Devpost submission deadline | |||
| 1:45 PM | Work time ends | ||
| 1:45 PM | T-Shirt Distribution | Klaus 1116 W | |
| 2:00 PM | Demos! | Klaus Atrium | |
| 4:15 PM | Closing ceremony & awards | CoC 016 |
