How to become a machine learning algorithm engineer (Zhang, the head of the referral algorithm department)
Being a qualified development engineer is not a simple matter, it requires a range of capabilities from development to debugging to optimization, and each of these capabilities requires sufficient effort and experience. To become a qualified machine Learning algorithm engineer (hereinafter referred to as algorithmic engineers) is even more difficult, because in the master of the general skills of engineers, but also need to master a small machine learning algorithm knowledge network. This article will be a qualified algorithm engineer to split the skills required to see what the need to master the skills to be a qualified algorithm engineer. Recommended Systems Engineer (Chen Kaijiang, technical CTO)
Now, even if "Big data", "AI", these words 360 degrees a day without dead ends bombing us, let us easily impetuous abnormal anxiety unbearable, but have to admit, this is as a recommended system engineer a good time. Recommended system engineers compared to normal-code farmers, the need for the PM to throw the requirements to the pixel level to achieve, so that heap code Chengshan; Compared with machine learning researchers, without indulging in mathematical deduction, to suppress a beautiful self-consistent model, unified academic debate; Compared to data analysts, there is no need to draw beautiful charts, Make cool PPT can give CEO report, embark on the peak of life. What is the position of the recommendation system engineer? Why do you need the skills mentioned earlier? The author will be combined with their own experience to answer. Dialogue Systems Engineer (Wu Jinlong, Einstein Interactive technology partner)
The dialogue System (dialogue robot) is essentially a machine-learning and artificial intelligence technology that allows machines to understand human language. It contains a variety of disciplinary methods for the integration of the use of artificial intelligence in the field of a technical focus of the drill camp. With the development of speech recognition, NLP and other technologies, the stage of dialogue robot will become bigger and larger with the advent of the interconnected age of all things. Data scientist (Lin Hui, DuPont business data scientist)
Before you answer this question, I'd like you to think about another question: why you should be a data scientist. Of course, if you are for 100,000 of dollars in the annual salary is understandable, but I sincerely hope that you will be able to associate this profession with your own value sense. Because it's hard to be a data scientist, but if you look at it as a way to achieve personal value, pursuing a goal can bring a lasting sense of accomplishment, and you'll feel happy and motivated in the process. Data scientists this position is also relatively new, so from the team building and career trajectory are still developing, with good prospects. I hope you can become a constant thinking, lifelong learning of data scientists. Heterogeneous Parallel Computing Engineer (Liu Wenzhi, head of High Performance computing Department of Shang Tang Technology)
With the hot of deep learning (artificial intelligence), heterogeneous parallel computing is more and more valued by the industry. From the beginning to talk about the depth of learning must talk about GPU, to talk about deep learning must talk about computational power. Computational power is related not only to the specific hardware, but also to the level (i.e., heterogeneous parallel computing) possessed by the person who can play the hardware capability. A simple analogy is: Two chip computing power is 10T and 20T respectively, a person's heterogeneous parallel computing capacity of 0.8, he got the computing power of 10T chip, and the heterogeneous parallel computing capacity of 0.4 of the people get the calculation of the power of 20T chip, and in fact, the final result may be a small difference between the two. The object of this paper is to tell the reader what knowledge is needed to become an engineer with strong heterogeneous parallel computing ability. Speech Recognition Engineer (Chen Xiaoliang, founder of Sound-intelligence technology)
At present, the accuracy and speed of speech recognition depends on the actual application environment, in the quiet environment, standard accent, common vocabulary of the speech recognition rate has exceeded 95%, fully reached the available state, this is the current speech recognition is more hot reason. The academic community has discussed many speech recognition technical trends, has two ideas is very worth paying attention, one is is the End-to-end speech recognition system, the other is G.E. Hinton recently put forward the capsule theory, the Hinton of the capsule theory of academic controversy is still relatively large, whether in the field of speech recognition is also worth exploring the advantages. In this paper, the main science, the knowledge of vertical and horizontal connection, and can combine practical and easy to understand the article, for a comprehensive understanding of speech recognition is very helpful. Seek technical breakthrough: the professional Path of depth study (Liu Xin, chief executive of China Branch)
Deep learning is essentially a deep artificial neural network, not an isolated technology, but a combination of mathematics, statistical machine learning, computer science and artificial neural networks. The understanding of deep learning can not be separated from the most basic mathematical analysis (higher mathematics), linear algebra, probability theory and convex optimization in undergraduate mathematics, and the mastery of deep learning technology can not be separated from the hands-on practice of programming as the core. There is no solid foundation for mathematics and computer support, in-depth learning of technical breakthroughs can only be castles in the castle. Therefore, it is necessary to understand the significance of these basic knowledge to the depth of learning in the beginners who want to have the achievement in the deep learning technology. In addition, our professional path will be from the structure and optimization of the theoretical dimension to introduce the depth of learning, and based on the depth of Learning framework practice analysis of the advanced path. This article will also share the experience of in-depth learning and gain in-depth learning of cutting-edge information. The actual path: Programmer's Machine learning Advanced Method (smart, Roullens Software co-founder)
If at school, we can not smell the opportunity of machine learning field, but choose other areas to learn and work ... Now I'm going to halfway decent and change machines to learn what to do to be as good as these people. Or, at least, good enough. The author's experience of painful transformation, say for reference.