For ambitious data scientists, how they stand out in the relevant scientific data with the job market? In 2019 there will be enough data science-related job? Or there may be shrinking? Next, let us analyze trends in data science, and explore how learning / AI field to get a good job in the future of big data and machine. "
1, you need a solid grasp of probability statistics, and to learn and master some algorithms, such as Naive Bayes, Gaussian mixture models, hidden Markov model, confusion matrix, ROC curve, P-Value and so on.
These algorithms must not only understand, but also know how they work. You need a solid grasp of gradient descent, convex optimization, pull on each methodology Lang, quadratic programming, partial differential equations, find the relevant product algorithm method.
If you are looking for a high-paying job, but also need to master the technology and machine learning algorithms, such as k-NN, naive Bayes, SVM and decision forests.
2、
Now most machine learning needs huge amounts of data, so you can not be a machine learning on a single machine. So, you need to use the cluster, you need to master a number of Apache Hadoop and cloud services such as Rackspace, Amazon EC2, Google Cloud Platform, OpenStack and Microsoft Azure and so on.
You also need to have a variety of Unix tools such as cat, grep, find, awk, sed, sort, cut, tr and so on. Because the machine learning are basically running on a Unix system, you need to master these tools, know their role and how to use them.
3, in the grasp of programming languages and algorithms at the same time, do not overlook the role of data visualization. If you can not allow yourself or others to understand the data, then they become meaningless. Data Visualization refers to how to present the data to the right people at the right time, so that they derive value. Primary data visualization tool comprising: Tableau, QlikView, Someka Heat Maps, FusionCharts, Sisense, Plotly, Highcharts, Datawrapper, D3.js, ggplot like.
4, to become data scientists do not have to get a degree in science data. In fact, you absolutely do not need to do this, but doing so is not a good idea. If you can get a computer degree, an engineering degree, economics degree, a degree in mathematics, statistics degree, actuaries degree, or a degree in finance degree in natural sciences (physical, chemical or biological) are possible. And even the humanities (including social sciences) is also possible.
Big Data to AI technology into the center stage of speculation, data science and machine learning began to emerge in all walks of life
Machine learning began to be applied to solve data analysis problems. Machine Learning, AI and predictive analysis has become a hot topic in 2017. We witnessed the data based on the value of innovation, including scientific data platform, the depth of learning and a few major vendors of machine learning cloud services, as well as machine intelligence, regulatory analysis, and behavioral analysis of things.
AI will now accelerate the pace of development, this year will be the rebirth of AI technology and scientific data to redefine the year. For ambitious data scientists, how they stand out in the relevant scientific data with the job market? In 2019 there will be enough data science-related job? Or there may be shrinking? Next, let us analyze trends in data science, and explore how learning / AI field to get a good job in the future of big data and machine. "
Enhance the technical strength
Programming languages and development tools
365 Data Science 1001 data gathered information from LinkedIn, scientists found that the greatest demand for the R language programming language, Python, and SQL. In addition, it Requires MATLAB, knowledge of Java, Scala and C / C ++ terms. In order to stand out, we need to master the Weka and NumPy such tools.
Probability statistics, applied mathematics and machine learning algorithms
You need a solid grasp of probability statistics, and to learn and master some algorithms, such as Naive Bayes, Gaussian mixture models, hidden Markov model, confusion matrix, ROC curve, P-Value and so on.
These algorithms must not only understand, but also know how they work. You need a solid grasp of gradient descent, convex optimization, pull on each methodology Lang, quadratic programming, partial differential equations, find the relevant product algorithm method.
If you are looking for a high-paying job, but also need to master the technology and machine learning algorithms, such as k-NN, naive Bayes, SVM and decision forests.
Distributed computing and Unix tools
现在大部分机器学习都需要海量数据，所以你无法在单台机器上进行机器学习。所以，你需要用到集群，需要掌握 Apache Hadoop 和一些云服务，如 Rackspace、Amazon EC2、Google Cloud Platform、OpenStack 和 Microsoft Azure 等。
你还需要掌握各种 Unix 工具，如 cat、grep、find、awk、sed、sort、cut、tr 等。因为机器学习基本上都是在 Unix 系统上运行的，所以需要掌握这些工具，知道它们的作用以及如何使用它们。
查询语言和 NoSQL 数据库
传统关系型数据库已经老去。除了 Hadoop 之外，你还需要掌握 SQL、Hive 和 Pig，以及 NoSQL 数据库，如 MongoDB、Casssandra、HBase。
基于 NoSQL 分布式数据库的基础设施已经成为大数据仓库的基础。原先在一个中心关系型数据库上需要 20 个小时才能处理完的任务，在一个大型的 Hadoop 集群上可能只需要 3 分钟时间。当然，你也可以使用 MapReduce、Cloudera、Tarn、PaaS、Chef、Flume 和 ABAP 这些工具。
数据可视化工具
在掌握编程语言和算法的同时，不要忽略了数据可视化的作用。如果无法让你自己或别人理解数据，那么它们就变得毫无意义。数据可视化就是指如何在正确的时间向正确的人展示数据，以便让他们从中获得价值。主要的数据可视化工具包括：Tableau、QlikView、Someka Heat Maps、FusionCharts、Sisense、Plotly、Highcharts、Datawrapper、D3.js、ggplot 等。
正确选择教育背景和专业
To become a data scientist, you do not have to get a degree in science data. In fact, you absolutely do not need to do this, but doing so is not a good idea. If you can get a computer degree, an engineering degree, economics degree, a degree in mathematics, statistics degree, actuaries degree, or a degree in finance degree in natural sciences (physical, chemical or biological) are possible. And even the humanities (including social sciences) is also possible.
But perhaps you'll get better development in other areas, such as economics, applied mathematics or engineering. First determine the scientific data this road is not suitable for them. 2019 will not let those interested in the field of data science a go of people disappointed. But then again, with a brain analysis capabilities, skilled programming skills, sincere enthusiasm and continuous self-improvement perseverance will determine the way your data scientists will go far.
View more articles:
I did not want to learn the basis of large data difficult?
Big Data learning portal, you have to master these skills
Three major technical direction of the large data field
How big data professional future employment prospects?
You teach compulsory three big data skills, quickly recorded