• data mining/machine learning/statistics
• data visualization
• database management
• programming

There are some free online courses that cover many of these areas, and these courses are usually part of a degree or a certificate program in data mining. Those who are new or interested in this field can learn a whole lot without paying a dime. Here is the list:

• Intro to Probability and Statistics (Carnegie Mellon)
• Machine Learning 101/102
• GovData (MIT/Harvard)
• STATS 120: Information Visualisation (The University of Auckland)
• R Programming (UCLA)
• CS 229: Machine Learning (Stanford) (videos)
• Linguistics 420: Statistical Natural Language Processing (Georgetown)
• SI 508: Networks: Theory and Application (University of Michigan)
• CS 591: Data Mining (West Virginia University)
• STATS 782: Computing for Statisticians (The University of Auckland)
• 6.867: Machine Learning (MIT)
• Andrew Moore’s Slides on Statistical Data Mining Tutorials
• Lots of tutorials (Data Mining Tools)
• Capstone project:  kaggle or kdd (for a bigger list see kdnuggets)

Some free text books:

• The Elements of Statistical Learning by Hastie, Tibshirani, and Friedman
• Mining of Massive Datasets by Rajaraman and Ullman

In addition, there is an excellent thread on quora on how to become a data scientist that covers lot of things and is a very good resource on the practice of analytics.

1. Used Yahoo Pipes (I created mine, but this one has more feeds)– this pipe aggregates feeds from different job web-sites, and gives the user unique job listing that you can subscribe via RSS: Job Feed Aggregator by Sean Dolan
2. Subscribed to the RSS feed for the keyword “data mining”
3. Copied the job descriptions and requirements of many jobs, and saved the text file
4. Got the python stemmer
5. Applied the python stemmer to the text file. Stemmer truncates words to their roots, so that we can combine variants of a word into a single word. (First or second step in text mining)
6. Created a tag cloud using the services of http://www.wordle.net/ . They use “stop words,” so I didn’t have to apply those. Stop words are common words, which necessarily don’t add any value for categorization, of a language.

Data Mining Jobs Tag Cloud

The most frequent word is: experience. Companies want people with experience in different data mining techniques. You’ll see that some other big words are: SAS (stemmed as sa), Excel, SQL, analytical skills, statistics, and quantitative skills.

And how do you master these skills, you ask?

1. Get a graduate degree in statistics, economics, mathematics, computer science, financial engineering, or industrial engineering with emphasis on databases, data mining, and marketing.
2. Successfully complete data mining projects using free, open-source data mining tools, such as Weka, R, Orange, Rapid-Miner.
3. Participate in data mining competitions. SAS’s data mining conference has a data mining competition every year.

Have a look at a detailed study by Pejic Bach, M: Creating profile of data mining specialist