Analytics Vidhya 2015 Year in Review: Best Content and Community Highlights

Introduction

What a remarkable year 2015 was for Analytics Vidhya! We not only made it past the dreaded "year 2" milestone where 90% of startups fail, but we thrived and grew faster than ever. When I left my cushy job at the end of 2014 to pursue my entrepreneurial dream, I wasn‘t sure how large the market for data science knowledge sharing could be. But I was passionate about filling that glaring knowledge gap.

Fast forward one year, and Analytics Vidhya has become one of the largest and fastest-growing data science communities in the world. Our web traffic increased by 500% in 2015 alone and shows no signs of slowing down.

In 2015, we launched several valuable new resources for our community:

• Discussion portal for Q&A and knowledge sharing
• Informative infographics and cheat sheets
• Salary test to benchmark compensation
• One-stop Resource Finder
• Hackathon platform to practice skills

Behind the scenes, our small but mighty team worked tirelessly to create and curate the best data science content out there. Our community clearly appreciated it, as evidenced by the wild popularity of certain articles, tutorials, and resources.

As we close the books on an epic 2015, let‘s look back on the most loved content of the year. Perhaps you‘ll discover an interesting article you missed or be inspired to test your skills in a different way. We hope it gets you excited for even bigger and better things to come from Analytics Vidhya in 2016!

Note: We always love hearing from our community. Feel free to share your thoughts, opinions, highlights and lowlights with us anytime at [email protected].

Top 10 Articles of 2015

1. Essentials of Machine Learning Algorithms

Choosing the right algorithm for a machine learning problem can be tough, especially for beginners. Should you use logistic regression or a decision tree? This epic post broke down 10 essential ML algorithms in both Python and R, complete with code. An awesome launching point for anyone looking to learn machine learning.

2. Top Data Scientists to Follow & Best Tutorials on Github

GitHub is an amazing resource for data scientists, not just web programmers. This article revealed some of the top data science experts to follow on GitHub and the best tutorials and code repos they‘ve shared. Learn from the best!

3. 7 Types of Regression Techniques You Should Know

There‘s a lot more to regression than just linear and logistic. This guide walked through 7 key types of regression analysis and how to apply them for predictive modeling. A must-read for both beginners and practiced data scientists looking to expand their toolkits.

4. Ultimate Guide to Data Exploration in Python using NumPy, Matplotlib, Pandas

Exploring and understanding a dataset is one of the most important skills for a data scientist. This super in-depth tutorial broke down data exploration and visualization in Python using popular libraries like NumPy, Matplotlib, Seaborn, and Pandas. Code included!

5. Difference between Machine Learning and Statistical Modeling

Machine learning and statistical modeling can seem quite similar on the surface. But while the lines between the two fields have blurred, there are still some key differences in approaches and techniques. This thoughtful post examined 7 important distinctions between ML and statistical modeling.

6. Top Youtube Videos on Machine Learning, Neural Network and Deep Learning

You might be surprised how many excellent machine learning tutorials and lecture series are available for free on YouTube. We compiled the best of the best into this must-watch playlist covering machine learning, neural networks, and deep learning. Perfect for visual learners or anyone looking for a lesson on the go.

7. Powerful Guide to Learn Random Forest in R and Python

Random forest is an immensely popular and powerful machine learning algorithm for both classification and regression problems. If it‘s not already part of your toolkit, it should be! Check out this beginner-friendly guide to getting started with random forests in R and Python.

8. Comprehensive Guide to Data Visualization in R

Data visualization is both an art and a science. While tools like Tableau and PowerBI make it easy to create basic charts and graphs, advanced visualizations require some programming skills. Luckily, R has robust visualization capabilities thanks to popular packages like ggplot2, leaflet, and lattice. Learn how to make a wide variety of charts, maps, and interactive visualizations in R with this guide.

9. Python Tutorials from PyCon Montreal 2015

PyCon is the largest annual gathering for the Python community. While the price of admission can be steep, many of the excellent talks and tutorials from PyCon eventually make their way online for anyone to access. We rounded up the most relevant and useful tutorials for data science from PyCon Montreal 2015 on topics like machine learning, NLP, data wrangling, and big data processing.

10. 9 Popular Ways to Perform Data Visualization in Python

Python has become the most popular programming language for data science, in no small part due to its impressive data visualization capabilities. Between Matplotlib, Seaborn, Plotly, Bokeh, Altair, and other libraries, Python can create almost any chart or plot you can imagine. See examples of 9 different essential chart types created with Python in this helpful roundup.

Top 5 Infographics of 2015

1. Cheatsheet – Machine Learning Algorithms

Machine learning algorithms are complex and can be hard to remember – especially when you‘re just getting started. That‘s why we created this super handy cheatsheet with Python and R codes for common supervised and unsupervised ML algorithms. Print it out, stick it on your wall, and impress your colleagues with how quickly you can whip up a model!

2. Job Comparison – Data Scientist vs Data Engineer vs Statistician

Data science teams have evolved to include a variety of specialized roles like data scientist, data engineer, machine learning engineer, and statistician. But what‘s the difference between them? We broke it down in this side-by-side comparison of key data science job roles, responsibilities, and skills.

3. 10 Movies Every Data Scientist Needs to Watch

Data science isn‘t all work and no play! Get inspired or learn something new by watching these 10 movies and documentaries with data, AI, and machine learning themes. From sci-fi flicks to true stories of data-driven triumph, these films will entertain and educate.

4. Essential R Packages for Data Science

The R ecosystem has thousands of packages for various data science tasks. But which ones should you focus on? We selected the top R packages for data import, wrangling, visualization, machine learning, and more. Use this handy map to navigate the most important R packages for each stage of a data science project.

5. 20 Must-Read Books to Boost Your Data Science Career

Books are still one of the best ways to learn a complex topic like data science cover to cover. But with so many titles available, where should you start? We picked 20 of the highest-rated and most recommend books on data science, machine learning, stats, and more. Build your data science library with this essential reading list.

Top 5 Discussions of 2015

Sometimes the best learning happens through discussion and debate with fellow data science enthusiasts. On our discussion portal, the community chatted about:

1. Comparing analytics programs and degrees – Which one is right for you?

2. The best statistics and analytics books – Build your data science bookshelf

3. Tableau vs QlikView – Choosing a data visualization platform

4. Creating word clouds in Python – A fun text mining and visualization project

5. Andrew Ng‘s Machine Learning course vs "Learning from Data" on EdX – Debating the best online programs for machine learning.

Join the discussion and tap into the collective knowledge of our community!

5 Tips from Top Analytics Vidhya Hackathon Winners

Practice is the key to improving your data science skills. In 2015, we launched our DataHack platform to host machine learning hackathons and modeling competitions. Hundreds of data scientists participated, but only a few came out on top.

We asked some of our hackathon winners to share their best tips for competitive machine learning:

1. Commit yourself to learning and mastering feature engineering.
2. Get creative with advanced techniques like H2O, XGBoost, and neural networks.
3. Ensemble methods are your best friend. Learn to combine and stack models effectively.
4. Hyperparameter tuning is both an art and a science. The only way to get good at it is through trial and error.
5. Don‘t get discouraged when your model accuracy plateaus. View it as an opportunity to learn even more.

Looking Ahead to 2016

2015 was a landmark year for Analytics Vidhya and our community. But we‘re just getting started! We have so many exciting things planned for 2016 and beyond:

• Expanded coverage of hot topics like deep learning, big data tools, IoT applications
• More hands-on tutorials, case studies, and projects to practice real skills
• Bigger and better DataHack competitions to really test your abilities
• Additional infographics, cheat sheets, and resources to reference
• Continued community growth and knowledge sharing

Of course, none of this would be possible without you – our incredible community of passionate data science learners and practitioners. Thank you for coming along on this journey and continuing to make Analytics Vidhya such an enriching and rewarding place.

We look forward to learning and growing with you in 2016 and the years to come. Onward and upward!