The Ultimate Data Science Learning Plan for 2017

With the explosive growth of data science, machine learning, and artificial intelligence in recent years, there‘s never been a better time to develop technical skills in these exciting fields. Whether you‘re an aspiring data scientist looking to break into the industry or an established practitioner seeking to expand your expertise, having a clear learning plan is critical for reaching your goals efficiently.

In this post, we‘ll lay out a comprehensive, step-by-step learning path to guide your data science education in 2017. We‘ll cover key topics and skills to learn, the best resources to use, and tips for applying your newfound knowledge on real projects. Our goal is for you to come away with a practical roadmap to follow in your journey to data science mastery.

While everyone‘s learning needs and backgrounds differ, we‘ll provide separate paths for three levels of students:

1. Beginners with minimal prior exposure to data science and programming
2. Transitioners from other technical or analytical fields looking to make a career shift
3. Intermediate and advanced practitioners seeking to deepen their expertise in specific areas

Feel free to jump to the section that best matches your situation, or read straight through for the full learning plan. Let‘s get started!

The Beginner‘s Path to Data Science

For those of you who are relatively new to data science, programming, and statistics – don‘t worry! Everyone starts somewhere. With the right resources, mindset, and a healthy dose of tenacity, you can develop a solid data science foundation this year. Here‘s our recommended roadmap:

Step 1: Ground Yourself in Basic Mathematics and Statistics (2-3 months)

Data science is an inherently mathematical and statistical field, so it‘s critical to develop a solid grasp of these fundamentals early on. We recommend starting with the following topics and resources:

• Linear Algebra: Basics of matrices, vectors, matrix multiplication, etc.
  Recommended Course: Coding the Matrix: Linear Algebra through Computer Science Applications (Coursera)
  Book: Introduction to Linear Algebra by Gilbert Strang
• Calculus: Key concepts like derivatives, integrals, optimization
  Course: Calculus One (Coursera)
  Book: Calculus by James Stewart
• Probability and Statistics: Probability theory, statistical distributions, hypothesis testing
  Courses: Introduction to Probability and Statistics (Stanford OpenEdX)
  Book: All of Statistics: A Concise Course in Statistical Inference by Larry A. Wasserman
• SQL: The lingua franca of data analysis and a prerequisite for advanced work
  Course: Intro to SQL for Data Science (DataCamp)
  Book: SQL Queries for Mere Mortals

Don‘t rush this foundational step – it‘s important to build your mathematical base before moving onto the sexier machine learning and analytics topics. Mastering the basics now will make the rest of your learning much smoother.

Step 2: Learn a Programming Language (Python and/or R) (2 months)

With some math and stats under your belt, you‘re ready to start actually working with data using a programming language. While there are many options including SAS, MATLAB, Java, etc., we recommend focusing your efforts on learning Python and/or R in depth. These open source languages have become the most popular tools in the data science community due to their versatility, ease of use, and rich ecosystem of powerful packages.

For Python, start with basic syntax and data structures and gradually work your way up to using popular data science libraries like NumPy, pandas, matplotlib, and scikit-learn. Some top resources include:

• Courses: DataCamp‘s Data Scientist with Python track, Coursera‘s Applied Data Science with Python Specialization
• Books: Python for Data Analysis by Wes McKinney, Python Data Science Handbook by Jake VanderPlas
• Docs: Python Data Science Handbook, pandas docs

For R, focus on data manipulation with dplyr, visualization with ggplot2, and machine learning with caret and tidymodels. Useful resources:

• Course: DataCamp‘s Data Scientist with R track
• Books: R for Data Science by Garrett Grolemund & Hadley Wickham, Applied Predictive Modeling by Max Kuhn
• Docs: tidyverse website, caret vignettes

As you‘re learning Python/R, start applying your skills on real datasets. Kaggle has an excellent repository of datasets to practice on. You can also find projects in DataCamp‘s practice mode and through R packages like fivethirtyeight and rcfss. The key is to get your hands dirty working with real, messy data!

Step 3: Dive Into Machine Learning and Data Analysis (3-4 months)

Now that you can work with data programmatically, you‘re ready to explore the world of machine learning and advanced analytics. Again, there are a ton of subtopics you could study, but we recommend starting with some of the most common and useful techniques:

• Regression: Linear regression, logistic regression, regularization
• Tree-based methods: Decision trees, random forests, gradient boosting
• Unsupervised learning: k-means clustering, hierarchical clustering, PCA
• Model evaluation: Train/test split, cross-validation, regularization, overfitting
• Natural language processing: Tokenization, TF-IDF, sentiment analysis, LDA, word2vec
• Big data: Spark, Hadoop, MapReduce, Hive

Some of our favorite machine learning resources include:

• Courses: Andrew Ng‘s Machine Learning (Coursera), fast.ai‘s Practical Deep Learning for Coders
• Books: An Introduction to Statistical Learning, The Elements of Statistical Learning, Hands-On Machine Learning with Scikit-Learn and TensorFlow
• Tutorials: Kaggle micro-courses, Machine Learning Mastery tutorials, Analytics Vidhya articles

At this stage, it‘s critical to practice implementing machine learning techniques yourself, not just reading about them conceptually. Tackle progressively more challenging datasets on Kaggle and other sources. Try replicating results from academic papers using open source data. The more hands-on experience you get, the better you‘ll understand the nuances of machine learning.

Step 4: Develop Your Data Science Portfolio and Personal Brand (Ongoing)

As an aspiring data scientist, one of the best ways to demonstrate your capabilities to potential employers is through a stellar portfolio of projects. As you‘re working through tutorials and implementing techniques on datasets, make your work public on GitHub. Write up your process and results in blogs and tutorials. Share your insights and visualizations on platforms like Kaggle and RPubs.

Establish your data science brand through social media and community engagement. Follow influential practitioners on Twitter and engage in discussions. Participate in local data science Meetups and hackathons. Contribute to open source projects related to data science tools and frameworks. Attend conferences like Strata, KDD, ODSC, and PyCon and learn from experts in the field.

The specific projects and activities you pursue will depend on your interests and career goals. The key is to demonstrate that you can extract insights from data and communicate them effectively. Establish an impressive online presence backed by a portfolio of interesting work.

Step 5: Prepare for Data Science Job Interviews (1-2 months)

As you start applying for data scientist positions, it‘s important to prepare for the interview process. Data science interviews tend to be highly technical, covering topics like algorithm implementation, statistical analysis, and open-ended analytical problem solving. Practice common interview questions on platforms like Interview Query and Eloquent JavaScript.

Additionally, brush up on your product sense and business acumen. Interviewers will often ask questions about how you would use data to solve a business problem or improve a product. Practice framing your answers in terms of delivering business value, not just technical accuracy.

If you‘ve made it this far and developed an impressive portfolio of data science projects, you should be well-equipped to ace the interview and land your first data scientist job! Continue honing your skills on the job and progressing in your learning journey.

[Transitioner and Intermediate/Advanced paths omitted for brevity… they would follow a similar structure focused on more advanced topics and a faster timeline]

Never Stop Learning

Becoming a data scientist is a continuous journey, not a finite destination. The field is evolving incredibly quickly, with new techniques, tools, and frameworks emerging all the time. Once you‘ve landed a data science job, it‘s critical to continue honing your skills and expanding your domain knowledge. Some strategies:

• Keep learning online through courses, tutorials, and books
• Attend conferences and Meetups to learn from other practitioners
• Read blogs and journals to stay up-to-date on the latest research and applications
• Pursue side projects to apply newly learned skills
• Contribute to open source projects related to data science tools
• Engage with the data science community online and in-person

Above all, stay curious and don‘t stop challenging yourself to tackle new problems with data. With the right learning approach and mindset, you can launch your data science career in 2017 and continue growing for years to come. Best of luck in your learning journey!