"MS in Data Science" the next MFE?

[Zeuge]

Seeing as data science is the next big growth industry, do you see masters' programs catering to the new profession popping up in the near future? Do you think that they might take the place of MFE programs as job prospects for MFE continue to decline?

 

[physEcon]

That's an interesting hypothesis. Hurry up and buy DataSciNet.com!

I don't think it will really happen since data science is more of a re-brand of statistics than a new field. An MS in Stats pretty much is an MS in Data Science (although I could be revealing my ignorance with this statement!). An FE degree made sense because there was a need to distinguish between really quantitative finance professionals and more traditional finance professionals who were better served by MBAs and MAs/MSs in Finance.

Another question that could help in this is: What was the big growth program before FE programs took off?

 

[Andy Nguyen]

Another question that could help in this is: What was the big growth program before FE programs took off?

Computer Science?
MFE is a niche degree that serves mainly the finance industry. When the industry shrinks or changes, it inevitably has affected the employment prospect of many MFE grads.

I don't know if they will come up with a new degree serving Big Data but CS depts would definitely take advantage of this and modify their curriculum to better meet the demand.

 

[Connor]

I've heard of a few MS degrees that are for business analytics and I see the trend continuing even though its basically just statistics and computer science. But I think at the end of the day, studying the foundations of computer science and statistics/math prepares you for this trend and any future.

 

[Andy Nguyen]

I know a few recent MFE grads took jobs as data scientist at Facebook, Google, LinkedIn, instead of financial firms.
I know more and more universities are offering MS degree in Data Science such as NYU, Columbia, Cornell, etc.
If you read the news lately, you sure heard of the new buzzword "big data" and its many applications in our life.
http://www.nytimes.com/compendium/collections/576/big_data
http://www.nytimes.com/2013/04/28/t...r-for-specialized-workers.html?pagewanted=all

Given the chance, would you do this degree instead of MFE?

 

[J Seo]

To be honest, as from Statistics background, Data Science is just fancy degree that includes Statistics + some Finance (or business). I think it's better off to go for MS or PhD in Applied Statistics (with heavy CS background) if one wants to be data scientist.

 

[Trev]

Data Science is just fancy degree that includes Statistics + some Finance (or business)

... and data collection, data munging, open source tools, cloud computing, visualization, machine learning, language processing, network analysis... Oh, and internships and career services. I'll take the fancy degree, and a few years back of my life, thanks.

 

[Trev]

I know a few recent MFE grads took jobs as data scientist at Facebook, Google, LinkedIn, instead of financial firms.

I've heard the same from some local MFEs. I've also heard rumbles about them asking for more exposure to Python & R in place of C++ & Matlab...

 

[J Seo]

Forgot to mention CS portion of Data Science. (n)
I took "Data Mining" graduate courses while I was in Bachelor's degree. and actually, I had fun. haha..and current research also involves big data.

As Connor mentioned, trend seems like converging to Statistics + CS. With heavy Statistics or CS background (or with years of professional experiences), Data Science degree will shape you better for Data Scientist. But without those background, this degree won't get you much just like MFE.

 

[Lyosha]

From reading about a lot of these "big data" etc. degrees/classes the only thing I can think of is that they are watered down programming classes that completely miss the point.

They teach you how to memorize steps to do the easy stuff, meanwhile completely skimping on the more important philosophical and creative underpinnings of the dark arts. And they produce heavy over-dependence on specific technologies, which are unlikely to be competitive or even maintained in the forseeable future.

In short, learn c++ and how the nuts and bolts of databases work. When I was working with "big data", understanding how my databases worked was much more instrumental than memorizing factoids about them. Because the challenges are dynamic and ever changing. Creativity goes a lot further than rigor, standardization and memorization.

 

[Daniel Duffy]

In other words, learn the trade and not just snippets. Learn something that can be used in 5, 10, 20, 30, 40 years.

I wrote my first semiconductor testers in 1974 using Basic on a teletype with paper tape LOL and since then the industry has shifted every 10-15 years. You need to be 2-3 years ahead of the decline of the current trends.

My tips (free advice)

. Learn functional programming
. Learn algorithms (esp. graphs)
. Parallel programming
. Statistics
. Numerical Analysis
. C++ (it will be here forever)

. Become a good programmer

. Learn 2-3 domains (domain knowledge). 'Big Data' means nothing unless you couple it to a domain.

These skills can be used all over the place.


// I have an interview with Andy a few years ago
https://www.quantnet.com/threads/interview-with-daniel-duffy.2699/

 

[Ujwal Gattupalli]

I am digging this old thread to see what is the current answer to the same question. A course in Data science and Business analytics is better than a MFE? The job options for DSBA are pretty wide where as the ones in MFE are tightly coupled to financial markets which are highly unpredictable. For a person from engineering background and programming work experience both the options are feasible but the questions is which is more beneficial.

 

[Phani Sarma]

From what i observe, the Data science is the latest "hot cake" being sold in the market and clearly the opportunities are far diverse than an MFE ( which traditionally caters to Finance audience ). That said, i would definitely re-iterate what Lyosha mentioned - the DS course these days tend to sell a technology stack to you - and without indepth knowledge of the underlying technology, the benefit from most such courses ( with superficial knowledge ) is going to be short lived.

It might boil down to these criteria -
1) Does Finance intrest you thoroughly ? then MFE.
2) Are you a Phd in Math , stats specifically - then DS. I believe Phd is added plus for MFE, but a Math Phd need not be like a pre-requisite. For DS, I think Math Phd should be must. If you are weak on Core DS technology Stack, and then just a grad in Math, you might not take off in DS.
3) Obviously the avenues for DS are better than MFE, and with the Silicon valley start up bubble, Data analytics is only going to go upwards for next 5 yrs.

Speaking for myself, i am completely for finance, hence shooting for MFE.I am from engineering + programming background, hence I am trying to learn DS basics - cos it gels well with applying the Python,R concepts.

 

[Dweeb]

Columbia’s Data Science Institute - Masters of Science in Data Science – coursework provides an in-depth overview of the subject matter.

 

[donny]

In short, learn c++ and how the nuts and bolts of databases work. When I was working with "big data", understanding how my databases worked was much more instrumental than memorizing factoids about them.

I'm somewhat agree with this statement, though it depends on the job trajectory of the person concerned. Indeed, every HF and bank needs a robust and proper data infrastructure set up. And the person who can do databases very well will always have that value to the company.

However, I believe its consensus that for this particular person, he'll be pigeonholed into just the database guy. It's unlikely the manager would want to switch the developer guy to the trading guy.

As I suspect those looking for the data science will want to eventually use the math and statistics they learn to analyze the markets, they probably don't want to be the guy, while valuable for database work, who knows just the nuts and bolts of the technical data infrastructure.