Data science is one of the hottest jobs in recent times, and often on the road will listen to passers-by about related employment issues. Data science sounds tall, and the skill requirements don’t seem to be as unattainable as CS. At the same time, with the US policy becoming more and more unfriendly to international employees, only a technical job can guarantee a bowl of rice, which caused Almost all international students to want to develop in this area. It is conceivable that the competition in this line will only become more and more intense in the future. Students who have the ability and background are still more secure to learn CS earlier.
Let’s talk about the employment of data science. “Data analysis” is an unambiguous position, and the United States is divided into many positions. Such as data engineer (more like code farmers, playing infrastructure), data scientist (skills requirements are also very high, everything must be understood), data analysis (relatively low-level, some entry, technical requirements are not high, comprehensive statistics, data and Business knowledge), business analysis (more on business analysis, such as how to reflect a business problem as data, reflected in the report).
First talk about what skills are needed to do data. Will grab data: to grab the original data from the web page, you need to use Regex, html, linux and other related knowledge. Will clean the data: R, python, SQL. Will store data: SQL, python. Will analyze data: statistics, machine learning, SAS, Numpy, pandas, Tensorflow. Will visualize the data: Tableau, matplotlib. Big data: Hadoop, Spark.
What ECON can help is basically to analyze this part of the data and a small part of the data visualization (R). Other knowledge needs to be supplemented later.
The following two levels are said: one is the relevant course that the NYU ECON project can provide, and the other is the course that individuals need to self-learn.
What class should I take?
Five compulsory courses: Statistics and Measurement I&II can help you learn the knowledge of statistics and data analysis, and it takes a little more time. At the same time, these are not enough. Further study on machine learning (NYU Courant), natural language processing (Stanford), advanced machine learning (NYU Courant) is required. Please consider the other three unrelated courses.
Five electives: Basic configuration considerations 2 NYU ECON + 3 Stern/Courant.
The College’s two electives recommend courses close to statistics. Economic and financial metrology is optional, but it is not recommended to choose a time series. The course is very general. The financial time series is good for teachers, but the course is average. Another option is the simpler course at the college, which spends more time on extracurricular learning.
In the three elective courses outside the college, it is recommended to choose Stern lessons in the first semester! Because: 1 may not be selected later. 2 Early school time is abundant. In this way, each course can be selected for an external course.
Recommendation: Stern Academy’s dealing with data course is recommended. The syllabus is set up reasonably. Although it does not teach statistical knowledge, it teaches how to capture data, how to process data, and the amount of data is quite large, and it has certain course difficulty.
Now the Stern college has already taught many data courses, such as data mining, marketing data analysis. If you want to choose, you can go to the syllabus first, or consult a professor to understand the main content of the course. Another point to note is that data analysis serves business analytics, so learning data under Stern can also learn a certain business analysis method.
Stern’s courses may require a prior course foundation, so first, check to see if you have a prerequisite course.
After learning these steps, you can supplement knowledge in data processing, databases, and big data. Other knowledge will be added to the next link.
Extracurricular course study
Stevens has business intelligence and business analysis project (which can complete a degree on the web), which has been around for a while and is relatively mature. One idea is to supplement the existing knowledge with reference to the course of this project. Another project that can be referenced is NYU’s Data Science project.
The Stevens project includes financial analysis and data management and data warehousing (data storage), data optimization process analysis, experimental design, database, and other courses are more important. The knowledge structure of this project is relatively comprehensive, and the knowledge requirements for doing a data analysis can be satisfied.
Better resources/courses: Stanford NLP on youtube, lynda.com/piaza, etc.
First, you can go to the Kaggle website to participate in the data contest.
Benefits: There is more experience in the data processing. If there is a ranking, you can also improve the quality of your resume.
Insufficient: The data in Kaggle is cleaned, and it is difficult to experience real data. The amount of data may be too small and many actual problems will not be encountered.
Conclusion: Finding relevant internships will help further.
The internship and work of the School of Economics are not very easy. But if there are a determination and goal to find an internship and work from the beginning, there is still a chance. It is not easy to find an internship in May, from the start of school in September to the second year of May.
There are still a lot of internship opportunities in New York. Pay more attention to it. Don’t spend too much time on the course to ensure that the grade point is too bad (teachers are good, not too bad). The pressure on the school is not very big
There are four ways to practice in the US (CPT, OPT, volunteer, blacksmith). Our project does not have CPT, you can use the project’s OPT, but the OPT will be reduced after graduation. A volunteer has no wages, and black workers are at risk.
For companies, especially large companies, especially want you to have a legal identity, so the probability of using OPT is still very large. The internship is still very helpful for finding a full-time job. The internship in the United States represents your skill level and your English ability is up to standard. Without an internship, the company will have certain concerns.
There are many ways to find an internship. The school will provide a lot of resources, and you should pay more attention to your job.
There are not many Chinese students who have found a job in the United States. Even if it is, it is just a small company. It can’t be a big or big investment bank. The difficulty of finding a job for a project depends to a certain extent on the efforts of the project’s director. Our current director is not very helpful in helping students find a job.
The old company uses SAS.
New company use: Python or R
Students in economics are more exposed to the problem of dealing with statistics, so R uses more. Small and medium-sized companies in the job market, new companies will use python more. Many large companies use SAS. Personal advice is that Python must understand, but also need to be able to implement various algorithms for machine learning. If you want to understand R, it is the same requirement, not only can you call the existing statistical package, but you can also write your own algorithm, such as gradient descent. Other required software includes SQL and big data related such as Spart, Hadoop.
In the process of looking for a job, I have a general understanding of the CS professional job search skills, or process, routine. Why is CS looking for a job, CS learning has a fixed routine, as long as you stick to it, you will have a high probability of finding a job, no matter what your major is, of course, this process is very painful.
But data is not the same as CS, and the technical requirements of data are not as high as CS (except data scientists). Data doesn’t have a complete routine, and no one can sum up a correct path to ensure that students can find a job. The market for data is not as big as it is supposed to be, and the competitive pressure is high, so it is necessary to be mentally prepared in advance.
There are many professional skills in the US professional market, especially Chinese students who don’t know about it. These skills are very smashed, but they are also very important.
There is a forum called an acre three-point land, where many people share their experiences in work, internships, and lottery. Participants in the forum will post internships and may even push in. The introversion is quite helpful for applying for an internship and finding a job. Many people on an acre of three-point land are very nice, because knowing that it is not easy to find a job, they will be willing to help the younger generation to find a job. You can contact them politely, leave contact information, or help.
There is a lot of experience in the details of the online application. For example, the success rate of online application is relatively low, but if you send a resume directly to HR, contact, self-recommended, the success rate will be higher.
In the year after graduation, you will run out of your one year OPT. Although there are two years of STEM extension, you must have an E-verified company to help you postpone, so you need to identify in advance and do a good job. ready.
If the OPT is used up, there is no signing, you can go to a school CPT, so the total is 4 years. If you can find a company that helps you pump H1B, it is great. The probability of a graduate student’s H1B is much higher than that of an undergraduate lottery. After pumping to H1B, it will be relatively stable, but it will be more troublesome when changing jobs.