Best Free Resources for Data Science for Beginners (pdf links available)

Courtesy DataQuest

Data Science is a broad term. It requires domain knowledge not only from Machine Learning and Data Analytics but also from Data Visualization, Databases(structured and unstructured), Statistics & Probability, Big Data, etc. So this time, I will highlight some of the best resources(according to me) out on the internet.

Starting of with the online courses to prefer, I believe practical implementation alongside theoretical knowledge is the best approach. Hence, You can try out with DataCamp or DataQuest where you can get all the requirements at one placed, arranged properly(I went for DataCamp).Heard of some fine reviews about IBM’s Data Science specialization course(Coursera) as well. Shifting focus, DataCamp’s Data Scientist Career Track has around 22 courses(seriously!!) covering the basics requirements that include-

1. Beginners Python/R(whichever track you take)
2. Data(frames) cleaning, manipulation & joining.
3. SQL (now a Mongodb course also available but not in the Career Track)
4. Data Visualization using matplotlib & Bokeh
5. Statistics basics
6. Supervised & Unsupervised Machine Learning
7. Introduction to Deep Learning using Keras

But as it is said, nothing can beat books when it comes to learning, so any online course should be accompanied by books as well. When it comes to books, I have strictly followed Oreilly media publications for this purpose. I will now cover each domain one by one for Python(not considering learning your preferred language’s basics)-

• Data Analytics- Python for Data Analysis it would be enough to start covering both NumPy and pandas quite thoroughly.
• Statistics & Probability- Think Stats, Think Bayes, Videos from Khan Academy & StatQuest (youtube)
• Databases-W3schools.com for SQL and if want to go for NoSQL, MongoDB: The Definitive Guide would be my first choice, though MongoDB documentation is also just awesome!!
• Data Visualization- Start off with Matplotlib, Seaborn (python libraries)and then can go for Tableau. For Tableau, Pluralsight’s/Udacity’s course can be taken up(you can do it on your own as well, just google).
• Machine Learning- Andrew NG’s courses are the best asset available. But if you wish to save some time or more interested in implementations, google up the basic algorithms and follow up documentation for sklearn. I myself learned it just reading some posts. Andrew’s explanations are also available at youtube for all courses.

I won’t recommend jumping to Deep Learning before practicing the above concepts thoroughly. Some of the best Data Science problems can be practiced at Kaggle, Analytics Vidhya & MachineHack. HackerEarth can also come handy for a few datathons.

Blogs from Medium, TowardsDataScience, Kaggle, KDnuggets & Analytics Vidhya can solve many of your problems during the learning period. Quora can be your life-saving guard. Stackoverflow remains the King!!

Though we have discussed online courses, books to read, domains to expertise, platforms to practice, blogs to follow up, but it all requires one’s interest as well to explore out things as there is no defined path for learning.

Explore more, Learn more