Flora
What do data scientists do?
What kind of companies do data scientists work for? How can I become one?
#data-science #data-analysis #big-data
9 answers
Kavita Jain
Kavita’s Answer
Hi Flora!
Data Scientist title is often used to describe jobs that vary drastically. Depending on the company and amount of data they have, it could mean pure Data Analysis where the person extracts data from a data store/warehouse and creates meaningful visualization and aggregation on top of that data. This branch answers questions around operational reporting of data.
It could on the other hand mean utilizing more sophisticated statistical and machine learning methodologies to extract patterns out of data and do predictive science based not that. This requires a formal degree in Maths, Statistics, Physics or a similar discipline. This branch is more focused on creating data driven products.
Hope this helps and God Luck!
Anumeha Goel
Anumeha’s Answer
Data scientist means a lot of things, but in general I can say the role includes a lot of working with data to do something useful. In some cases, such as in my job, it might be to answer questions nobody else has answers to, in business speak referred to as "drawing insights". In some cases, the purpose is to build algorithms to predict things. I notice you have posted a few other questions about finance, and I think in this field the predictive analytics might be most relevant.
The basic requirement for a data scientist is to be good at quantitative analysis. Make sure you are taking math and economics classes, and it might even be helpful to learn some programming software such as Stata, R, SPSS, mySQL, Python etc.
Matt Brown
Matt’s Answer
The definition of this newfangled title "data scientist" is basically a super-awesome statistician, with a variety of skills and experience to match (I see the title as a senior title, and have seen it at the C-level). Since you can apply statistics to a wide variety of information, this title would be transportable across a variety of industries where extracting information from large amount of structured or unstructured data (aka "big data") is useful, such as pharmaceutical, network security, defense/intelligence, medical, financial, government...
If you're interested in quantitative analysis of the financial markets, my suggestion is to start working on a model. In fact, a financial quant I used to work with spent his free time working on a football (aka soccer) player performance prediction system, while I (as a network engineer), spent very little time looking at utilizing simple stats to extract data out of network traffic. You can use these examples of your expertise during an interview.
Thanks for the helpful explanation and advice, Matt!
Flora
Yi Xu
Yi’s Answer
we use data science to understand what drives people to do something online, from clicking an ads on the website to make a purchase, and use that conclusion to optimize how we allocate resource
Dai Li
Dai’s Answer
Data scientists in a product-oriented organization (think about companies like Airbnb/Uber/Facebook) are there to build better products. They make it happen through 1) aiding better decision making 2) participating in building data-powered product. For 1) data scientists perform quantitative analyses, build dashboards and run experiments. For 2) data scientists design and implement algorithm-based, automated solutions that are typically consumer facing.
Yi Xu
Yi’s Answer
you will either be working almost as an engineer to implement the AI algorithms, most likely work for cool start ups , such as companies want to develop self-driving car, or developing a device that can detect cancer from early stage etc
or as a AI researcher in a lab - these types of jobs are very limited and hard to find. you can either work for a school or some of the leading tech companies will have their own research center such as google brain @google
Yubing Sun
Yubing’s Answer
In terms of the work scope, data scientists cover quite a bit from business understanding, data understanding, data ETL, cleansing, screening, feature engineering, modeling, validating, test design and analysis, and feedback loop. Then it will goes to maintenance and such analysis work on an ongoing basis.
Team work is also very important to really address each step/area of data science and drive the effort to a success at the end.
Patrick Meyer
Patrick’s Answer
If you're considering a career as a data scientist, there are several key steps you can follow. First, build a strong educational foundation in subjects like mathematics, statistics, and computer science. Next, develop technical skills in areas such as programming and data handling. You might also want to consider further education to specialize in a particular area.
Building a solid portfolio through practical projects is a great way to showcase your skills. Participating in networking events can open doors to new opportunities. Gaining hands-on experience through internships is invaluable, and keeping up-to-date with industry trends will keep your knowledge fresh. Lastly, honing soft skills like communication and problem-solving will make you a well-rounded professional.
By following these steps, you'll be well-prepared for a successful career in the ever-evolving field of data science.
Sumant Sahoo
Sumant’s Answer
In my view, Data scientist job is beyond big data/data mining/statistics/programming skills.
A data scientist should be these qualities to be successful in any industry:
- Business Acumen - This is most important to understand the business problem before you jump into modelling/machine learning tasks.
- Communication - A good data scientist who can convert the business problem to statistical problem and similarly who can convert statistical solution to business solution. Basically you should be able to communicate in business language to business guys and technical language to technical teams.
- Analytical techniques - it can be start with descriptive statistics to prescriptive to predictive. I think most of them covered in earlier answers
- Tools or programming skills - this also covered in earlier answers; R, Python and Big data technologies would be helpful
Hope this clarifies your question.