Skip to content

hyonschu/DataScienceResources

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

48 Commits
 
 

Repository files navigation

##Data Science Resources

This repo is intended to provide open source resources to facilitate learning or to point practicing/aspiring data scientists in the right direction. It also exists so that I can keep track of resources that are/were helpful to me and hopefully for you.

I aim to cover the full spectrum of data science and to hopefully include topics of data science that aren't either actively covered or easy to find in the open-source world. For instance, I haven't focused on in-depth machine learning theory since that is well covered. If you are looking for ML theory I would look to some of the online courses, books or bootcamps. There is a lot of theory information available online, some is linked lower on this page here, here and other info is available with many purchasable books.

Keep in mind that this is a constant work in progress. If you have anything to add, any feedback, or would like to be a contributor - please reach out. If there are any mistakes or typos, be patient with me, but please let me know.

Lastly, I would add that a large portion of data science is exploratory data analysis and properly cleaning your data to implement the tools and theory necessary to solve the problem at hand. For each problem there are many different ways and tools to execute a successful solution - if one method isn't working re-evaluate, re-work the problem, try another approach and/or reach out to the community for support. Good luck and I hope this repo helpful!

#Table Of Contents

  1. Data Science Getting Started
  1. Data Pipeline & Tools
  1. Product
  1. Career Resources
  1. Open Source Data Science Resources
  1. About Me

Section of the data pipeline & resources:

Data Science Getting Started

Data Science is a multidisciplinary field covering at the very minimum - statistics, programming, machine learning Drew Conway's venn diagram. These topics are covered throughout this repo. I personally find the best way to learn a topic is to get my hands dirty quickly - with that in mind I would probably get to work in python and then implement different tools or theory into my toolkit as I understand each element. If you haven't used python before I would strongly urge you to use the codecademy course to familiarize yourself with the content and how to program. Good luck and have fun.

Starting

Data Science Courses:

  • Coursera - Data Science Specialization at Coursera - many other courses available as well.
  • Udacity - Online MOOCs that are the Data Science related courses.
  • Data Science Bootcamps - List of all bootcamps currently on the market as of April 5, 2014.
  • Coursera Machine Learning Course - Coursera pinnacle Machine Learning course.
  • Edx - EDX courses related to data science.

Data Pipeline & Tools

###Python Python is my workhorse language specifically as it has many data science and statistic library, the ability to work in production environments, and work on other problems outside of data science. There are many other languages that could be useful but are not covered here: Julia, R, Cython, Pig, Scala, Java, etc.

####Stats/Engineering Libraries A collection of workhorse libraries that are elemental for any python data scientist.

####Data Acquisition Libraries that are very helpful for abstracting away some of the complications of scraping or working with HTTP.

####Processing & Exploratory Data Analysis A collection of documents explaining some of the ways to do processing & EDA.

###Databases/Frameworks A collection of databases & frameworks that are helpful for data management and are the industry standard.

###Machine Learning There is a lot of information available online about the theory, mathematical intuition, tuning for this discipline. I am not trying to cover it in that depth, at least not at this current time. These are some high level knowledge posts and toolkits.

####Model Selection Resources about how to decide on your model.

####Model Evaluation Resources to help with understanding model evaluation.

####Feature Engineering A critical element of Data Science to improve your performance but minimally talked about.

Additional Tools or Processes

Resources on other topics that are very helpful for data scientists and product.

Data Visualization

Collection of the best libraries that I know for easy and powerful data visualizations.

  • ggplot - ggplot for python ported by the team at yhat.
  • matplotlib - Awesome plotting library for python.
  • d3 - De facto gold standard for polished visualization - in js, steep learning curve but beautiful outcomes.
  • bokeh - Interactive visualization library.
  • d3py - Another library for data viz.
  • vincent - Help with python for d3.
  • seaborn - Clean statistical data visualization library.

Other available Visualization Resources.

  • Scott Murray's D3 Tutorials Tutorials from Interactive Data Visualization for the Web
  • tributary.io - live code visualization platform designed specifically for D3.js
  • plot.ly - A web visualization and data processing platform
  • blockspring - Share code and visualizations through a single platform
  • dot.append - Ian Johnson (enjalot) goes through several live-coding examples using D3

Design Theory

The importance of design theory in data visualization and presentations could not be understated. Through better understanding of design theory and principles, a data scientist can convey more information and meaning in their presentations.

Ipython Notebook Tutorials

Collection of ipython notebooks that are helpful as examples to either using tools or to explain certain topics.

Data Sources

Collection of sites to access data if you want to build out a project or just use some of the tools for EDA.

New Data Tools

Aim to keep track of developing trends and new tech that is helpful for the practicing Data Scientist. New might be a misnomer.

  • BigML - machine learning for the everyday user, also useful for EDA.
  • GraphLab - graph-based, high performance, distributed computation framework. They just implemented deep learning onto their platform.
  • ModeAnalytics - platform to share analysis/data science.
  • Apache Mahout - Scalable machine learning library. Not in python.
  • Apache Hadoop - Open-source software for reliable, scalable, distributed computing.

Product

###Product Metrics
Understanding product, user behavior, and product metrics is helpful for data scientists in industry. Being able to help your product manager and team execute on strategies by understanding the problem, metrics and what they understand facilitates a more fruitful relationship.

Team Communication & Business Tools

There are some very innovative new companies that are producing very effective tools to minimize and abstract away inefficient processes at companies. While it isn't strictly data science related, these products could be very help to integrate with your teams to improve overall productivity.

  • Aha! - Clean product roadmapping software for PMs.
  • Slack - Amazing team communication tool - abstracting away unnecessary e-mails.
  • Harvest - Effortless time tracking for business.
  • Trello - Helping organize everything - great for project management.
  • Zapier - Bringing together Harvest + Slack + Trello and a lot more...
  • Thoughtbot Playbook - A detailed account of how thought book runs is software consulting company talking about guiding principles, design sprints, code reviews to sales and operations. A content packed post.
  • IFTTT - 'Putting the internet to work for you'. Great for small companies to automate social media, marketing or to have your own personal recipes set up.
  • Github - Clearly a great product - 'Build software better, together'.
  • Web Analytics & Reporting Software:
    • Google Analytics - In depth real-time analytics.
    • Mixpanel - provides real-time analytics and solid cohort analysis.
    • Clicky - Pride themselves on ease of use.

Career Resources

Data Science Career Path

Types of Data Scientists

Not all Data Scientists are the same and it's critical for organizations to understand what it is they need, and how best to fill those roles and/or complement the skills of their team. Finding the organizational structure that enables the data scientists/data engineers within the organization and generates better results is also crucial. It should be given thorough consideration.

Data Science Applications/Use Cases

Data Science has so many different applications and use cases within industry - many are continuously discovered. These resources provide some potential ideas.

Data Science Websites/Books

More resources for community based information or hard copy books.

  • Data Science Handbook - Not yet released but should be interesting providing stories from academia and industry about data science - go read the post for a better description!
  • CrossValidated - A question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization.
  • StackOverflow - Language-independent collaboratively edited question and answer site for programmers.
  • Kaggle - Model building competition and great resources for training and data.
  • O'Reilly Media - A lot of content rich books available and tutorials on using the tools.
  • Quora - Question and answer site - lots of data science content and career content.

Data Science Meetups in the Bay Area

A great way to meet other Data Scientists and keep up to date with best practices.

Data Science Blogs

The name say's it all.

Design Blogs

Data Science Conferences

  • Strata - Conference and a lot of videos from previous conferences - great resource.
  • GraphLab - Another great conference.

Data Science Presentations

Relevant Business Processes

  • Lean Startup - A method to develop product and businesses.
  • Agile Development - group of software development methods to optimize for self-organizational and cross-functional teams.
  • Scrum - an interative and incremental agile software development framework for managing product development.

##Open Source Data Science Resources While the name might sound redundant this section represents other sites or repos that have aggregated information covering similar topics. Tons of great content on these sites - definitely go check them out.

Other Open Source Data Science Content

There are some really great resources linked within this section covering all of Data Science, the entire data pipeline, machine-learning, statistics, python, etc. Go check them out.

Auxiliary Content & Apps

ABOUT ME

I acquired my skills through programming in an on-the-job environment and then taking three months off to learn and put into practice my data science skills @ Zipfian Academy. For me taking that time off to learn, run the daily/weekly sprints, and be in a collective learning environment at Zipfian was irreplaceable. Even if Zipfian resources were open source, without taking the time off work and having the drive to learn all the necessary material would be next to impossible. I am always interested to hear what other data scientists are doing and using for tools. I am interested in a wide range of different open source &/or private projects - feel free to reach out on Twitter @sf_oak or LinkedIn. Or go check out my start-up venture capitalist recommender ~ finding the long-tail of the VC community at findryouvc.co.

Analytics

About

Open Source Data Science Resources.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published