Selecting a Data Science Masters
Вставка
- Опубліковано 14 жов 2024
- Jump to the good stuff: 4:09
Data science masters degrees are popping up all over the place. This somewhat long video points out a variety of features that separate the good and not so good programs. There are literally over 100 programs claiming to be data science or machine learning related. Many have the title of business analytics however many of these are not really data science programs.
Financial Engineering: CDS Are Dying:
www.quantnet.c...
Why You're Not a Quant
• Why You're Not a Quant
Data Science vs Big Data vs Data Analytics
• Data Science vs Big Da...
Forbes Article
www.forbes.com...
Data
www.analyticsv...
www.mastersinda...
SUPPORT THE CHANNEL
Quant t-shirts, mugs, and hoodies:
teespring.com/...
Connect with me:
/ dimitri-b. .
/ dimitribianco
☕ Show Your Support and Buy Me a Coffee ☕
ko-fi.com/fanc...
Very useful, thanks a lot! Would appreciate it if you could make a similar video for an MS in Statistics (top programs, cost, compensation, duration, caveats i.e. good vs not-so-good programs, job opportunities, US unis vs UK Singapore, MS vs PhD etc.). Based on your risk management (which, if I understand correctly, requires a strong stats background) experience and your affinity for statistics, I feel that such a discussion would be useful.
Excellent vide as always, very much in agreement with this video, except that I do not see how much value it is to be in a classroom instead of online because all the reasons given can be done online in most cases, so the cost of Moving to a classroom does not compensate for the supposed value contributed. I think this has to do with people not yet getting used to it.
As for business, remember that it is where many of us want to apply analytics, it is important to know the terrain on which one works.
'Very Good Delivery'! Your research is superb in giving an understanding of the 'Master in Data Science' not generally given by counselors...thank you very much...
Thanks
You just made me rethink of applying to Msc Analytics program at UChicago. Debt part is real, it takes atleast 4 years to clear the student debt. You always give a description that is more prudent. I appreciate words coming out of real quant people like you.
Thanks
Hi Dimitri. First of all, I've been watching your channel for awhile. I was also in your situation too. I was admitted to one of those top 10 ranked MFE program on Quantnet in Fall of 2017 and dropped out within 2 months. My big turn off was how students were only interested in money. They didn't know who Edward Thorp was. I also didn't like the program director. That person was too full of him/herself. While I was preparing for MFE application, I took PhD level classes in applied mathematics at UC Berkeley - Numerical Methods for PDEs, Numerical Linear Algebra, Theoretical PDEs, and Stochastic Process. So when I was in MFE program, I didn't really feel motivated to study when classes were relatively easy compared to what I took at Berkeley. So I wholeheartedly agree with you, when you said you don't like the interdisciplinary degree. I dislike it too. I took a year and half off, and spent my time at Stanford and Berkeley, taking as many PhD level classes in machine learning and statistics (and CS) as I can. That's when I realized machine learning is really statistics. Just like how reinforcement learning is originated from control theory and is rebranded as reinforcement learning.. I also agree with you that I'm very skeptical of these data science program popping up in the US... Just like MFE program was popping up in the US years ago. The reasons why I'm skeptical of these data science master program are (1) you can't really call yourself data scientists when you don't know statistics. (2) students are trained how to use libraries, but have no ideas how to reproduce research papers from the scratch - this is important, because you will realize how many research papers published in the top peer - reviewed journals like the Nature, NeurIPS, JMLR, etc are just false ( as John Ioannidis at Stanford said most research finding is false in 2005) . (3) data science tries to find intersection among mathematics, computer science, and statistics... which make these students jack of all trades, master of none. (4) there is a bubble within deep learning. Many practitioners within deep learning community are even not admitting the existence of deep learning bubble. All these autonomous vehicle, personalized medicine, drug discovery via deep learnings are just hype. I know that people like David Donoho at Stanford and Michael Jordan at Berkeley are openly calling out bullshit within deep learning. Sorry, it went a little bit longer than I expected.. I watched your videos several years ago and I was happy to find you online. I know how hard it must have been dropping out MFE program. Lastly, most of these students are drawn by $. I can't really blame them as I'm also partially driven by $. Thanks for pointing out that machine learning is really just statistics. I frankly don't know what deep learning is.. the more I study, the more I find it as an alchemy.. Last year, there was a researcher from Cal who gave presentation at NeurIPS that deep learning is an alchemy and it drew criticisms from the deep learning practitioners - a big red flag when they can't handle constructive criticism. I'll stop here.. Thanks again Dimitri and hope you have a great weekend !
I really appreciate the feedback. The longer I've been in the industry and the longer I make videos the more I feel like I'm crazy. There really aren't too many people who love to learn any more. Everyone is looking for a shortcut. The most common questions on my channel are based on, "what's the bare minimum to be a "quant"/ "data scientist" and make millions?" When I mention we don't get paid millions or you have to learn your entire life they get mad, call me names, and/or just leave.
The data science/AI/machine learning non-sense is really interesting. To start with everyone has a different definition of the terms and if you don't use their terms then you are just dumb and clearly not a data scientist or whatever title they are claiming. To me (like you) it's just statistics and math applied in a different way. It scares me to see people thinking it's a magic solution that will fix the world. I explained the math of an MLP neural net to a senior person at a bank and they didn't believe me because it was too simple. There had to be some secret that would just blow their mind since everyone has been talking about the ability of these to solve every problem. I find the area interesting and am excited to add a few new methods to my tool belt. (I've also made a few videos on the over hype and lack of actual evidence but I've never published them due to the backlash I would get from the community).
Keep up the hard work of learning and making a living. I too enjoy the money but definitely enjoy the work more than the money which is what keeps me in my current industry and job. Again thanks for being a part of the channel!
@@DimitriBianco Hi Dimitri. Thanks for your thoughtful feedback. I like your channel because your opinion is genuine. Believe it or not, your video on dropping out MFE program helped me to follow my decisions. It was risky move at that time for me (my program director called me I'm crazy that I'll regret this decision... ). I think you also mentioned in the past about your skepticism on ranking of MFE programs by Quantnet. I agree with you wholeheartedly. Actually, I was obsessed with ranking of schools many years ago; however, ranking is deeply flawed (with seriously questionable methodology like you mentioned. ) . For instance, NYU medical school was never "top school" but for some reason, this year it is now at the same level as UCSF, Harvard and Stanford medical schools, due to bogus methodology. I think ranking by QuantNet is deeply flawed, to put it mildly.
You're not alone in thinking that you feel like you're the crazy one. I lived in NYC and now live in the Bay Area, and people I see everyday are machine learning researchers at say Google Brain, Facebook AI Research (FAIR), OpenAI, and many sketchy startups that sell themselves as "AI". And I feel I'm the crazy one because I see the dangers these "researchers" hype their researches and overpromise to the public so that they can command 6-7 figures salary. Did you know that newly minted PhD graduates with PhD in machine learning can get 7 figures salary? It's quite common and I think this is one of the big reason why traditional quant MFE programs like Berkeley's MFE and some in Manhattan are sending more students to these tech companies, instead of financial companies ( not realizing that only PhD can get that 7 figures). I get it. Financial industry is on a secular decline since its peak in 2007-8 crisis. But, these MFE programs need to make some serious (or complete ) adjustments to its curriculum, if they want to train their students as data scientists as well as quants. Otherwise, (to me) they are lying to their students that they are MFE programs and yet want to sell their students as data scientists.
I still wouldn't recommend undergraduates to go for master programs. It takes time to learn materials and schools like Stanford uploads most of materials in machine learning freely available. Why would you want to spend 1-2 years and $100,000 or more, when you can learn machine learning / data science / deep learning / stat for free ? Going to master program is not the only way to connect with other people / practitioners in the field.
If they want to gain theoretical insights, I would rather recommend them to go for PhD. It's also quite sad, because the minimum qualification will be so skewed upward that people without PhD in machine learning, statistics, CS, whatever can't even get an entry level job at this rate.
Yeah, almost all companies (especially after Google made nice job hyping up Alpha Go success - which is not even really novel, and tech companies are all in arms race among others to get all the talents) foolishly believe machine learning will solve all these problems, be it sluggish revenue growth, thin profit margins, etc. Your supervisor probably doesn't know that machine learning is really statistics and there is no barrier to entry to deep learning like NLP (unlike quants, where you still need some decent backgrounds in mathematics to understand PDEs, Stochastic DEs, numerical methods). If you look at CS224 (NLP) at Stanford , prerequisite is almost nonexistent. Your supervisor foolishly believed that only machine learning experts can explain NLP, not quants. It happened to me too, because I took CS224 and thought this is really just alchemy.
If you are interested in these fields (I think you're since you use statistics every day), this link hdsr.mitpress.mit.edu/?Twitter&HDSR_launch&HDSR contains some commentaries by famous statisticians on data science in general. It is under "Michael Jordan - AI Revolution Hasn't Happened Yet." You can read commentaries from Jordan, Donoho, Candes, etc. I have taken classes under all of them them and they are far more critical on deep learning in lectures. David Donoho was the chief scientist at Renaissance Technology. This blog post also provides critical commentaries on deep learning ( blog.piekniewski.info/ ).
I watched your recent video on applying machine learning to finance and agree with you. We're dealing with totally different problems in finance (where data being scarce, signal to noise ratio being high, and not being stationary distribution) compared to technology ( where you have billions of data, etc). Even if your deep learning / machine learning algorithms are brittle and useless beyond the training sets (most of the time, these algorithms are useless.. just look at self- driving cars), the failures are not mission critical. If Google / Facebook serves me with wrong and totally irrelevant ads, it has no impact on me. However, in finance, people lose money. In medicine or self-driving car, people die.
Whenever I see someone who says machine learning will solve all the problems in ______ ( ex. self-driving, drug discovery, personalized insurance, or more recently investment strategies), I get very suspicious. For example, you mentioned this guy in one of your newer video, Marcos Lopez de Prado kept saying on his twitter that "machine learning will solve the problem in investment strategies", "econometric is useless", "If Francis Galton, Karl Pearson, Ronald Fisher and Jerzy Neyman had had access to computers, they would have created an entirely different field: Machine Learning." That last quote troubles me, because any person who takes CS229 at Stanford ( Machine Learning class) will realize that ML is really statistics that CS people rebranded as machine learning. That and his self-promotion of his book and research papers are deeply concerning to me..
Lastly, on people who want to take short cut, these people will be the first to get fired when another recession takes place. If they are in 20s and still believe that there is a short cut to the summit, I don't feel sympathy for them when they lose their jobs. I would rather have people like you telling me the truth, rather than those program directors of MFE programs and their students lying to prospective students.
Sorry, it got really out of control.. It was really nice seeing that there are people like you. I know you won't but don't change the direction and tone of your channel, just because some people can't handle the truth and constructive criticisms. Thanks Dimitri, and I wish I have met you in my cohort so that at least I didn't feel like I wasted several years of my life preparing for MFE program to realize that most students have zero interest in learning financial engineering ( and care only about making quick $). How can they not know Edward Thorp, if they want to become quants ? I don't get it.
@@christiansong227 how can I reach out to you directly? Would love to chat more on this discussion and pick your brain
I applaud you for giving your uncensored opinion on the matter. Yes data science is just statistics rebranded. And if you want to do something mathematical within your job, you are practically obliged to go into the data science field in our country. There isnot much else left to do here. Also the reason why I will have a second masters degree in statistics next to my masters in physics and astronomy.
My eyes turned in light when I saw this video just uploaded. Say that you made it for me😍 thanks Dimitri!
Andres Rossi I actually decided to make it after a conversation with Andy Nguyen from quantNet
Hahaha well, of course I'm not the only one in the world 😁 btw, do you think a master degree in statistical analysis (which is made for data analysts) should be good to work as data scientist or in the future will be needed a specific data science master degree?
As of right now I think the degree title is less important as long as you get the skills required to do the jobs you want in the future. As of now there are many names for data science related programs and I think these names are unreliable. Many low end programs will add degree names that look great however the programs won't deliver the skills needed in the job market. So yes, I think a masters in statistical analysis is good for data science as long as it covers a lot of the main areas of stats and computer programming used for data science.
Thanks to you I'm having more clear ideas about these work fields. I guess statistical analysis is good to work as Quant analyst or risk manager as well, while data science is a lot more specific. By the way I have big issues with UA-cam notifications, I just saw your comment casually
Your videos always speak the truth. Seriously its laughable when people think that data science or ml is a new degree.. Also everyone these days are data scientists, data gurus... Considering my own gruelling experience with econometrics its laughable how superficial their claims are. Most data scientists just mug up tools without delving into the engine or mathematics that drives these models. Personally i think this mushrooming of data science courses without proper mathematical or statistical models are creating a lot of nonsense inferences
**It's just statistics** .... legendary comment...
To answer your other question on where to find the best data science schools, I would look for schools with strong statistics departments.
Hey Dimitri, really love your videos, especially this one. I am very torn between a masters in data science/applied stats and mathematical finance. I have a solid quantitative background but I am pretty new to both fields and I'm in my 3rd undergrad year right now so I'm mostly researching programs and trying to schedule visits to some of these schools and find out more from the directors themselves. Some of the programs I'm really interested in are NYU DS, UWash Statistics, Rutgers Math Finance and most particularly Stevens Institute of Technology Financial Analytics which is all Data Science classes but specifically applied to Financial Markets. I'm curious as to what you think of it. Keep up the great videos!
This all depends on what you want to do as a career. If you want to work in risk management I would focus on statistics. If you want to work with derivative products I would focus on financial engineering. If you want to be an equity analyst (non-quant) I would focus on financial analytics. Data science is a hot topic right now however I don't think it will be a career in finance. Banks are looking to hire a few right now however in my opinion it's just another stats tool. It's like getting a degree in linear regression. If you want to only do data science then I would keep in mind you might get a job in finance however you'll most likely end up in tech or business analytics.
@@DimitriBianco Ok thanks for the information, I will keep all of this in mind. I was under the impression that DS masters and Financial Engineering masters had more overlap the difference being that DS is skewed more towards stats/cs and financial engineering is skewed towards math.
Thanks Dimitri
Happy to help. Thanks for watching my videos!
Du what is your master degree?
Great Video Dimitri! What do you think about the Data Science course offered by LMU Germany? And also what are your thoughts about doing Data Science in European university?
I took a look at their program and it looks really good. I like the focus on stats and info science as well as it's 2 years with a capstone project. From what I can see, it looks really good.
Not sure if I'm behind on your videos (you may have covered), but if not, could you do a video on relevant coding languages to get in risk mgmt or quant work in general? I know it's also dependent on the company one is at as well other factors, but a generalized review would be beneficial. For example, is R a dying language in the professional field, only to be utilized widely in academia due to it's long hold?
I can make a video about this. To give you a short answer, almost everyone is moving to Python.
@@DimitriBianco appreciate the prompt response. I figured that, but have seen job postings that requires c++, R, and SAS, or matlab. The first two in my opinion are archaic in the working professional world, the latter two are more frequent in academia from my understanding. Thanks again!
Thanks for the info, any thoughts on UC Berkeley's Data science program? I really liked their curriculum but as a newbie to this it's hard for me to tell how it is seen in the industry.
From my perspective it doesn't look very good. One red flag is that it's online. A lot of companies avoid online degrees regardless of the school name. The second issue I see is that they write a lot of fluff about how great the program is but there aren't classes explicitly listed on the curriculum page. I would look for an in person data science masters.
Interesting video Dimtri! I am curious about the US' IEOR master programs that offering concentration in analytics, would you consider it as good pathway to be a real data scientist? Also, how would you consider taking MA in statistics as a pathway to study machine learning?, knowing those machine learing stuffs are actually statistics.
Thankss Dimitri!
Getting a masters in statistics would be, in my opinion, the best route to a data science / machine learning job. From industry experience the IEOR programs seem to lack rigor.
Hi there! Thanks so much for this video. I was wondering what your thoughts were on the Masters of Data Science at UBC in Vancouver?
It's hard to tell from reviewing websites an curriculum as universities can list anything. One way to decide on a program is to ask them (or search on LinkedIn) to see where they are placing students. If the jobs look interesting then it would be a good program for that career.
can you please give me the list of programs you divided into stats and math and business and also the electives list and there breakdown I'm really confused with all the names for all the degrees and there programs & courses that they offers really want to focus on getting the actual knowledge for becoming a scientist in data science.
What do do you think about the Master of Statistics: Data Science program from UW-Madison?
From an overview of the courses, it looks like a good program. The one thing that stuck out as a not a good thing was that they allow you to take undergrad courses. If you selected the right courses though, it could be a strong degree.
Great video! Very informative. I know you prioritize skills in statistics, but what is your take on getting an MS in Computer Science, particularly from an engineering school? The goal would be to create data science tools or create models and put them into production. Personally, I think this may be beneficial in that it will focus more on application than theory if let's say another MS CS degree was under a mathematics department. (I'm specifically looking at the NYU Tandon MS in CS)
I agree with you and think CS can add a lot of value. NYU's program looks very good and you are right that it is more applied. A lot of data science is about optimization which is best for CS majors.
Dimitri Bianco I also agree with how you said that business knowledge can be gained on the job. And I think technical knowledge might be better gained from academia. Would you rank a data scientists skill set to be stat > CS > business? Although of course all would be important depending on the responsibilities.
Ismael Cruz that's a tough one. I think I would have stats=CS > business. Modern stats is heavily focused on programming and data science has become more resource intensive which is why I think stats and CS would be equal.
I am graduating this fall and have recently decided to pursue a master in data science. I live in the Florida I was looking at UCF MS in Data Analytics. The program looks appealing. My one caveat is that they still have open application, ending dec 1, for the spring which is odd since most other schools applications ended in October and September. Any thoughts are greatly appreciated
In general the programs looks good. My only advice would be to try and focus on a specific area or industry. I'm guessing the reason they have a later date for applications is because almost every school has some sort of data analytics program which has made this area more competitive for students.
Great Video Dimitri! can you please guide me between Masters in Data Science vs Masters in Machine Learning.
Thank you
The difference is really just marketing. I would consider both degrees but focus on what courses they are offering. A good foundation in statistics and math will better prepare you for a job. There are some good programs out there that blend the statistics and math into data science/machine learning courses.
I think to some degree these schools market the idea of data science. What’s wrong with an MS in Statistics ? Someone who knew the stats and math could apply it wherever
what's wrong with an MS in Statistics: Statistics sounds old, boring & dusty;
Data Science sounds fresh, edgy, exciting & promising
Yea but still when you're talking about data science you're still talking about statistics and computer engineering. It's just a name, not a skill, as Dimitri rightly said too. To make it easier, 2+2 will always be 4, so statistics will always be actual science
A lot of my colleagues in risk management have MS or PhD in Statistics. Many of us are now learning and using data science in our daily work. An MS in Stats is good but as Eryk points out, some employers might want a flashy degree name. Personally I'd rather have someone with a Stats degree than a data science degree.
Hi Dimitri,
Could you also talk about the top tier Phd programs in Finance, Statistics, and Financial Economics / Econometrics? In particular, which are the top 3 schools in each fields, what are some key differences in the core courses, hot research topics with quantitative finance / risk management applications, job prospects (both academic and industry), and general recommendations for someone who is looking to apply to one of these programs.
Here is a brief background of myself. I have a BSc in Statistics and a Master in Financial Economics. I have worked in various roles in the Actuarial Department of a global insurance company (Valuation, Asset Liability Management & Investment Strategy), during which I also finished all the actuarial exams (Corporate Finance and Enterprise Risk Management specialization). Afterwards, I had an opportunity to run a manufacturing & trading business due to family reasons.
Now that I have worked in a global corporation and ran my own business, I believe I'm ready to take the next step in my career. I have always enjoyed learning, and it is always my goal to do a Phd. My dream is to become a professor in a well-respected business school and do applied research and perhaps consulting work on the side.
Given my background, I am interested in finance or risk management. But I’m having difficulties deciding between the 3 fields - Finance, Statistics, and Financial Economics / Econometrics. I came across your UA-cam channel while researching about them. I really enjoy your videos, especially the one where you talked about the difference between a quant and an actuary.
I am look forward to your reply. Any comments / advice are welcomed!
I will try to address some of your industry questions in a few videos in the future but for career advice for you...get a PhD in Finance. Business schools respect finance degrees however they look down on economics and statistics (at least in the US). Business professors make a lot more money than quantitative fields as well. The average statistics professor might make $85,000 while a finance professor might make $200,000. Overall there is a disliking between business programs and other programs including economics. Personally I think PhDs in finance lack rigor compared to other PhDs such as statistics and econometrics.
Take a look at MIT, Chicago, and Stanford for a finance PhD.
Dimitri Bianco, thank you so much for your quick reply. Your advice is certainly helpful - to the point and makes a lot of sense.
I look forward to watching more of your videos. Keep up the good work!
hi dimitri thanks for the video, so informative.
by the way i am about to take a master degree in data science of business at university of stirling in scotland. here are some of the courses that are available in the program :
- leadership
- professional development
- representing and manipulating data
- business analytics
- statistic with R
- data analytics
- business consulting group project
electives (20 credits/1-2 courses):
- risk management
- strategic management
- project management
- innovation management
- relational databases
- mobile financial applications
the duration is 1 year.
my question is, is it good enough for me to compete in the future after graduate?
and do you think 1 year is an ideal time to finish this degree?
thank you so much.
Hey Dimitri what do you think about a masters in statistics from USF in Tampa, they have a very robust 10 class masters program focused on pure and applied Statistics.
Maybe l am wrong, but l think it is better to get a degree in the quantitative hard stuff and learn the easy (business, data analysis, etc) stuff on your own .
I don't know a lot about USF's program but a statistics masters does provide a solid foundation. It opens a lot of different opportunities in the analytics space.
@@DimitriBianco yeah l am not a big fan of the Data science programs, l prefer something generic and work the specifics on my own. I just wanted to have your take.
Thanks
Hey Dmitri, thanks a million for the video. You threw light on a lot of important aspects which no one does on UA-cam. I have recently been accepted for my Masters in DS at Technical University of Dortmund. I know you are kind of based in US, but it would be really helpful for me if I can get your opinion on the program ? :) TIA !
The core courses look fairly good. Like most masters, choosing electives that have a common focus will really help for finding a job.
Hey Dimitri,
Are you familiar with the University of Southern California’s MS in Applied Data Science?
I was recently accepted, but it’s very new and I’m worried that it might not focus enough on statistics.
Any thoughts?
Thanks!
USC is a good school and I like that the program is in the Engineering school. My only concern is that some of the classes such as INF510 assume you have no background in programming. I would assume all of the materials to be at a masters level. The program could be good given the university's name and hopefully the higher classes have more rigor since they are in the Engineering school.
Thank you for the reply, Dimitri! Yes, unlike the MSCS Data Science, this program is more geared toward students with limited programming experience. I graduated from a small school where I double majored in Math and Economics. I’ve only taken introduction courses in Maxima and Python thus far, so this could be a good fit.
Any thoughts on the Computational and Data Science MS from Chapman University in Orange, CA?
Kyle D It depends on what you want to do after graduating. My opinion is that the program looks to scattered and lacks depth. I'd ask for their job placement rate and industries for those placements. At the end of the day we're all looking for a career after the degree.
I live in the Denver area there really aren't many programs, there is Regis and University of Denver . University of Denver hasa good program but is very expensive. Regis program doesn't seem very good.any thoughts
Hey Dimitri just curious because I am interested in pursuing a masters in statistics but can't decide which one is right for me. Would you recommend someone to pursue a traditional pure MS Statistics program as opposed to the following options such as an Applied Statistics, Data Science or Risk Management/ Financial Statistics degree program? Which degree you think is more advantageous or appeals to you the most?
That's a very challenging question to answer. A few factors to consider are the school's reputation and topics covered. I like to see programs that have a focus which indicates you covered fewer topics at a greater depth. The schools reputation helps with finding a job and can indicate program quality. A lot of companies like to hire people from local schools as well, so the location can also be important. I will try to make a more detailed video on this topic.
Hey Dimitri, I'm a junior Finance and Business data analytics double major at ASU and looking to get an internship and eventually a job at a finance firm. I have about a 3.6 gpa and am wondering if there is anything else I can do to boost my chances at landing a good internship. Ideally I would like to get into equity research or something similar.
You should spent your time on writing a good resume and practice interviewing. One of the biggest mistakes I see is students trying to write the resume as if they have work experience. When companies hire students (and interns) they want a really good student. If they wanted a really experienced professional, they would find someone who is already working.
Also make sure your resume has a focus. You might be good at many things however a company is usually hiring for a very specific set of skills not a jack of all trades.
Hey Dimitri,
Great Video..Can you Please help me out choose universities in Germany as I'm really confused which one to choose.
TU Munich has two courses: Data Engineering and Analytics & Maths in Data Science
LMU Munich, University of Manheim, RWTH Aachen University also has courses on Data Science.
I'm finding it really difficult to choose among them.
Please help me out.
Thanks in Advance.
There are a lot of things to consider however in general I would apply to a few programs as you might only get accepted to one program. I'm not in Germany or Europe so you'll need to use your judgement given your perspective however I will give you some points to consider.
1) Europe/Germany job markets: Look online to see if there are a lot of job postings with "data science," "data engineering," or "data analytics" as the job title or skills required. If a lot of jobs list data science instead of data engineering then you'll know it will be easier to get a job if you have data science listed in your degree title and class selections.
2) University reputations: I would consider the reputations of the different universities. Getting a job after school will be related to the school's reputation. If the school is well known it will be easier to get a job and network.
3) Course topics: I would select a program that focuses towards your interests. Look through the curriculum of the different programs and figure out which one looks the most interesting. If you find the material interesting you'll do better in school as well as in the industry.
If you are wanting to do data science, I would encourage you to make sure the program consists of math and statistics. There are a lot of business programs world wide offering data science degrees however you will not get the same rigor as a computer science, engineering, or math department. I took a quick look at TU Munich and they look like a solid university and good program. I didn't look at the other programs in Germany however I think TU Munich is a good measuring stick for other programs. I like that they have three areas of study so you can specialize in what interests you. Choosing between data engineering and data science will depend on what you want to do. Data engineers focus more on data infrastructure where as data scientists focus on extracting information from the data. The blog below has a better description.
blog.panoply.io/what-is-the-difference-between-a-data-engineer-and-a-data-scientist
@@DimitriBianco Thanks a lot Dimitri. Your tips really helped. Cheers!!
what are well rounded data Science programs that are offered by some university, that actually focus on becoming a data scientist with knowledge on core subject like stats and cs also lill bit of (around 10%) business knowledge say electives that are offered in business.i would like you to give some examples of universities (names) so i can look for myself how they are better. your response would be of great value to me.
Email me (see response to on of your other comments) and I can give details on my list which mainly focused on more well rounded top programs.
Hey Dimitri I have a question.
I currently work as a data analyst and I really want to move towards data engineering. I was thinking of doing master in data analyst from Wgu or bachelors in comp science from Wgu.
Would you recommend this school for anything?
In general I don't recommend online degrees. It makes getting a job very challenging because you're competing with a lot of other students from well known brick and mortar universities.
Hey Dimitri? What's your take on UVA's latest data science offering?
dsi.virginia.edu/degrees/info/programs-and-courses
I would love to know your thoughts.
If I did decide on this I would be an in state student.
I'd say it's above average for the MSDS.
Pros: UVA has a good reputation as a general university, the cost would be cheap for you, there are 2 capstone projects, and they have a good selection of electives.
Cons: The program is only 11 months compared to other programs that are 2 years. It's really hard to get a deep understanding of any of the concepts without spending more time learning.
You might see if you can find their job placement rate and the average starting salary. These will give you an idea of your return on investment.
@@DimitriBianco Thanks mate,
I think it is a very new program so they may not have a lot of post graduation data out there yet. I will continue to look into it though.
Wow I think I haven't learned so much for a while.
Hello Dimitri,
Really loved this video of yours. Thanks for posting the same.
I recently got admitted into the UChicago's MSc Analytics program. Please share your thoughts/insights about this program. Thank You.
It really depends on your end goal. If you want to work doing business analytics, I think the program looks fairly good but it's hard to tell from just a website. If you are looking to be more specialized as a data scientist, I would ask the program to see how many data scientists they have placed. The program looks to cover a lot of topics however I would wonder how deep each class covers a specific topic.
Hello Dmitri. First of all, great video. very informative. Would you please give me an evaluation of Data Science program at Higher School of Economics, Moscow. I recently got accepted there. it is time for me to make a decision. I would really appreciate your input. Thanks in advance
Hi Dimitri, What are your thoughts on MSc Data Science from The University of Sheffield?
www.sheffield.ac.uk/is/pgt/courses/ds
The program looks about average. A lot of the courses look to be introduction courses though. Stronger programs will have more classes that are deeply focused.
The website below should help in thinking about how to select a data science program.
www.analyticsvidhya.com/blog/2016/07/10-analytics-data-science-top-universities-masters-usa/
How About going with M.S in C.S With M.L/ A.I track after have had 2 years work ex in business analytics.
That could be a good route.
@@DimitriBianco thanks for the reply mate
Firstly, thanks for this context! I’m gonna apply to this program and can you tell me how it look from content? www.uni-potsdam.de/en/studium/what-to-study/master/masters-courses-from-a-to-z/data-science-master.html
It's hard to tell given that the class titles are so general. It's hard to tell what you would cover and at what depth. Overall I think the program looks fairly good however I would select a specialty and try and take as many classes in one area. Being a generalist always makes finding a job hard.
Hey Dimitri, what do you think of studying the teachings of Warren Buffett and other great investors? I’ve found it far more practical and useful than finance and business classes.
Lj H finance and business class provide a good base but they do have a poor understanding of the complex markets of today (I'm agreeing with you on this). Studying famous investors might provide some good lessons however most of the famous investors leverage their political power and amount of capital (Buffett is a perfect example of this). To really understand markets these days a solid background in finance and quant finance is required.
Hi, what do you think about the curriculum of Master of Science in Applied Data Science at Syracuse university?
It looks okay but I don't like that it's online and the course topics look too general for both the core and the electives. If I was hiring for a data scientist I most likely wouldn't hire a Syracuse Master of Science in Applied Data Science.
Dimitri Bianco i am not talking about the online course. I mean the on campus program. Also if I end up taking some subjects related to neural networks would it help me in getting a better job?
This really depends what you want to do. Their program is very qualitative (not quantitative). By this I mean the programs looks to focus on using programs like R to plot data, build simple models, and organize data. If you want a job that focuses on business analytics this could be a good fit for you. If you are wanting to build machine learning and AI models, this program won't give you the knowledge (math and stats) required.
If you do want to do machine learning at a very detailed level I would look for programs which have courses such as:
Linear Algebra and Matrix Analysis
Machine Learning and Predictive Analytics
Probability Theory
Algorithms for Data Science
Deep Reinforcement Learning
Dimitri Bianco Thank you
How about leuphana(lüneburg),management&data science vs. hildesheim, data analytics?
It's hard for me to judge because I don't know the reputation of these school but I did search for their curriculum to see if I could compare the course topics. I couldn't find any specific class list for Lüneburg but for Hildesheim they did provide the curriculum. For Hildesheim, I like that it's two years long menaing you will learn quite a bit. The general classes seem to provide a good overview which can indicate a lack of specialization however they have good specialized tracks for electives and I really like that they lab courses and a thesis. The thesis and labs is where you will really learn a lot of details and become more of an expert than a general program that only has classes. Overall from what I can see Hildesheim looks like a really good program.
thanks for your reply. this is leuphana`s program: www.leuphana.de/fileadmin/user_upload/grad_school/files/2Master/Studiengangsflyer/EN_Management_Data_Science.pdf
I chose leuphana because I thought ıt ıs a better start for busıness fıeld. It has less theoretical courses but more business ones which can give enough flexibility to select a fitting area.
I have 3.6 CGPA in my bachelors, Could you recommend some top schools in UK where i can get admission. Thanks
I'm not sure given I'm not from the UK. Maybe someone else on this channel could provide some insight.
What about Stanford data science?
Hey interesting video. Hoewever, next video consider speaking a bit more softly, not so loud and intense. It gets quite hard and annoyting to keep on listening. Maybe your microphone is not good. Keep it up though:)
I want to study economics for undergradutate degree but for master degree ı want to study data science.Can ı get master degree on data science?
Yes this is a possible route.
Part time master programs are ideal just pretty much voided this video.
How is Masters Data science in RMIT
how is master of data science at University Of Auckland
It's very hard to tell from a website. It looks to be more of an applied degree than a rigorous degree. I'm guessing their students go more into business analytics however it really depends on what you want to do. I would email the program if I were you to see where they are placing most of their students. If the jobs sound interesting then I would apply.
I'll give you my mail id if you agree to send the data and i really appreciate all your research your video is jus one of the kind that is super helpful it's very clear about a lot of stuff that has made this data science feild a very confusing thing for people who aren't specifically from both stat and computer background.i personally am from computer science background and i want to get a job as a data scientist because that's what i like and obviously money too but really get the actual shit when I'm spending all the money is important so that's why I'm asking you for the breakdown of the pie charts and i hope u read this comment.