This course teaches data analytics techniques and how to discover and exploit patterns in data, and how to mine and analyze extremely large data sets.
Increasingly, technology companies are applying data analytics techniques to masses of public and private data to identify unusual events, recognize patterns, and make predictions. Sources of such data can include the World-Wide Web, social media, and increasing, devices such as smart phones, appliances, vehicles, electric meters, et cetera – the "Internet of Things." The ability to deal with data of these types and at this scale will prove to be a high-demand skill for data analysts as applications of commercial interest increasingly go beyond business intelligence.
In the first course, Understanding Customers, students will work as data analysts for a simulated company, Blackwell Electronics. The students' job is to use data mining and machine-learning techniques to investigate the patterns in Blackwell's sales data and provide insights into customer buying trends and preferences. The inferences students draw from the patterns in the data will help the business make data-driven decisions about sales and marketing activities.
Students will use the open source WEKA machine learning package to understand the relationship between customer demographics and purchasing behavior. Next they will use WEKA algorithms to develop a “recommender system” to determine which add-on products a customer will be likely to buy. Finally, students will present to management, explaining their insights and suggestions for data mining process improvements.
After gaining experience with key statistical machine learning techniques and with delivering business value through analytics, students will move on to Predicting Profitability and Customer Preferences, where the students' job is to extend Blackwell's application of data mining methods to develop predictive models.
In this portion, students will use the R statistical programming language augmented with machine learning packages to predict which potential new products that the sales team is considering adding to Blackwell's current product mix will be the most profitable. Next, students will create a model to predict which brands of computer products Blackwell customers prefer based on customer demographics collected from a marketing survey. Finally, students will again present to management, explaining their insights and suggestions for data mining process improvements.
Note that although these projects are set in a retail context, the real focus of Data Analytics – Discovering and Exploiting Patterns in Data is on acquiring facility with a range of machine learning techniques that are applicable in a wide range of domains.
The third course, Web Mining, will coach you on how to mine and analyze extremely large data sets to gain insight into real-world business problems. Students will conduct sentiment analysis utilizing cloud-based computing and machine learning tools, and will interpret the results to make and communicate predictions of vital interest to business stakeholders.
Students will be working as data analysts for a simulated data analytics consulting firm, Alert Analytics. Their client, Helio, developed a suite of smart phone medical apps for use by aid workers in developing countries. The government agency will be providing workers with technical support services, but they need to limit the support to a single model of smart phone and operating system. Helio has created a short list of devices that are all capable of executing the app suite's functions. To narrow this list down to one device, Helio has engaged Alert Analytics to conduct a broad-based web sentiment analysis to gain insight into users' attitudes towards the devices. The students' job is to conduct this analysis.
Students will use the Amazon Web Services (AWS) Elastic Map Reduce (EMR) platform to run a series of Hadoop streaming jobs and the WEKA machine learning package to develop a predictive model that will infer user sentiment towards devices from the lexical content of web pages extracted from the Common Crawl of the World-Wide Web.
In the final course, Deep Analytics and Visualization, students will be working for an Internet of Things technology start-up that wants to use Data Analytics to solve two difficult problems in the physical world:
Smart energy usage: Modeling patterns of energy usage by time of day and day of the year in a typical residence whose electrical system is monitored by multiple sub-meters.
Indoor locationing: Determining a person's physical position in a multi-building indoor space using wifi fingerprinting.
Students will learn to use the R statistical programming language to perform visualizations, then to generate descriptive statistics and predictive models using time series regression techniques and statistical classifiers. Finally, students will present the results to the start-up's management, explaining strengths and weaknesses of the approaches that were implemented and making suggestions for further improvement.
Students can complete the program in either 22 weeks working 30 hours per week or in 44 weeks working 15 hours per week.
This program is designed for professionals who want to acquire new skills in the areas of data analytics and big data. You’ll learn how to conduct analyses of data that yield business value, to interpret the results of your analyses to make predictions, and to communicate data mining results to management and other non-technical audiences.
Upon completion of this program, you will be able to:
Identify types of business problems for which data analysis can provide significant insights in support of business decision-making.
Translate business objectives into data mining opportunities.
Install, run, and apply statistical machine learning tools to different kinds of data.
Apply data mining in retail business and other contexts. Through these activities, you will become broadly competent in the use of statistical machine learning techniques of classification and regression.
Interpret the results of data analysis to make predictions and to establish the reliability of those predictions.
- Communicate data mining results to management and other non-technical audiences.
Because the program is online, it is open to people around the world. When applying, you should have:
At least a year of work experience
Familiarity with Windows, Mac, or Linux operating system, specifically:
Creating and managing folders within folders
Creating and extracting files from zip archives
Elementary administrative tasks (e.g., installing software requiring admin privileges)
Basic familiarity with Microsoft Office or an equivalent productivity suite
Basic knowledge of statistics may accelerate your initial progress in the program, but all necessary statistical concepts will be introduced during each course.
- Understanding Customers
- Predicting Profitability and Customer Preferences
- Web Mining
- Deep Analytics and Visualization
A note from Team XTOL:
The next offering of Data Analytics/Big Data Bootcamp will take place 04/03/2017 from 9:00am. The class will meet for sessions which will be held in Online. We hope you can join us. If you can't make this session, check below for more offerings.