more info/ register
Use actual industry-specific data sets, and solve problems with them
I'M a QUANT (math / stats professional)
I'm a coder (Computer professional)
I'm a business professional
You know it and we know it. Eigen systems-based dimensional reduction has been hanging around since the 1930s (via Hoetelling (1933) and Thurstone (1935)). Supervised MVA (via Fisher's multiple discriminant analysis (1936)), and backpropogation neural nets since Werbos (1975). Yet, the modern era of data science algorithmics didn't really kick in until the late 1980s -- when CNN's via backpropogation trained by gradient descent (Waibel (1987, 1989)) began to appear. Then....boom! The golden era of ML arrived, and with it, the explosion of supervised machine learning, SVM, ensemble methods, baggers and boosters, and all the rest. I know. This is likely old-hat to you. If not, we will cover it in the bootcamp. The question, though, is why should YOU -- you mongo quant you -- sign up to THIS bootcamp? Two reasons. First, you won't waste your time. Second, you will get to play with some really advanced stuff. See below!
IF YOU ARE A REAL QUANT, MANY DATA SCIENCE BOOTCAMPS WILL WASTE YOUR TIME. NOT THIS ONE
The primary educator in this course, Dr. James A. Libby, is a pioneering marketing scientist, data scientist and strategy consultant. Dr. Libby's data-science based protocols and decision support platforms have been deployed in 60 major strategy engagements for 20 F500 companies in 40 countries. Jim has been writing algorithms and software that have been presented at major marketing science conferences since the early 1990s. Jim's Ph.D. dissertation applied (there is that word again) computational linguistics to disentangle long-standing issues of authorship in ancient documents using multivariate visualization, a variety of non parametric modeling (HLLA) approaches and information theory. What all this means is that this bootcamp will stress the applied side of data science -- as well as speaking to coding issues. What else is on the menu? In terms of algorithms, we will cover the usual suspects in clustering, data visualization, SVM's, ensemble methods, T-SNLE, etc., but we will do something more. We will describe how automated goal-solving methods which converge upon optimal hold-back validated outcomes form the foundational basis for decision support (DSup). DSup is the end game of the people that pay your salary and mine. For example, to see how DSup works in real live business terms click on the decision support tools related to Innovation/Positioning, Product Design, Customer Experience/Sat/Loyalty, Predictive Analytics, and Market Segementation. That should get your quant juices flowing!
PLAY TIME. WITH TOYS YOU CAN TAKE HOME
You know this history too. Data science is more than integrating data across the organization and applying super-massive knowledge discovery, predictive analytics and cloud computing. It's folding in automation to get to answers to business questions quicker. And getting to those answers faster than the next firm -- without ceding power and flexibility to the big box vendors. The Judson Data Science Bootcamp. Solving REAL business problems with REAL data. And that's the point, isn't it?
You know it and we know it. In the 1990s machine learning began its golden era -- and with it emerged the modern era of data science coding. But the data science coding world has undergone dramatic shifts since the 1990s! While C++, C# and Java still have their place, two tsunamies flooded through the world of data science coding:
THE OPEN SOURCE REVOLUTION: R, Python:
Remember the 1990s? OOP was all the rage. C++, and C# reigned. If you needed an algorithm you had to buy an algorithm library. How gauche! Then came R and the open source revolution that put power in the hands of the coding proletariat -- from which all true revolutions come. If you are a coder, we expect that you may know R or Python. Even if not, our goal will be to apply and automate automate these language with repositories and frameworks, so you become the guru in your organization.
THE DISTRIBUTED DATA AND COMPUTING REVOLUTIONS: HADOOP AND CLOUD COMPUTING:
You know this history too. The way forward is clear: integrating data across the organization and applying super-massive knowledge discovery, predictive analytics and cloud AND local computing. But you won't merely be using Hadoop and Cloud Computing here, you will be using it to solve REAL business problems with REAL data. And that's the point, right?
You know it and we know it. Data science is powerful. But it must always, always solve a real-world business issue. That's why we began our design of the Judson Data Science bootcamp by performing a national competitive, risk and demand assessment. Here is what we found:
THERE ARE LOTS OF DATA SCIENCE BOOTCAMPS OUT THERE. CURRENTLY 29 IN FACT. BUT THERE ARE VIRTUALLY NO FULLY APPLIED BOOT CAMPS
Here is a graphic to help make our point. As of late 2018, there were 11 major types of bootcamps in the competitive marketplace:
Did you notice that there were no bootcamps that start with the discrete business problem to be solved?
The Judson Data Science Bootcamp is different. From end to end, it is an APPLIED data science bootcamp. An applied bootcamp has three advantages 1) It allows us to teach data science inductively. Inductive learning actually accelerates learning and retention. 2) It allows us to compare and contrast different coding, math, and data approaches to inductively make an argument for best practices. 3) It allows us to solve problems in "pods" -- groups of learners who solve a problem just like they must in the "real world." To see the seven discrete problems we use as modules to teach data science click here.
THERE ARE LOTS OF ACADEMIC PROGRAMS IN DATA SCIENCE (CERTIFICATES, MASTER'S DEGREES, AND DOCTORATES) OUT THERE. BUT IN ADDITION TO ACADEMIC EXCELLENCE, WE NEED A PATH TO CONNECT THE ACADEMIC CONTENT TO THE BUSINESS PROBLEM
You probably know this history too. Data science is more than integrating data across the organization and applying super-massive knowledge discovery, predictive analytics and cloud computing. It's folding in automation to get to business answers quicker -- and getting to those answers faster than the next guy without ceding both power and flexibility to the big box vendors. It's solving REAL business problems with REAL data. And that's the point, right?
Nullam vestibulum pretium quam, a