IBM China Research Institute, SAP, Netease game, and IBM Xi'an CDL test interview questions

IBM China Research Institute, SAP, Netease game, and IBM Xi'an CDL test interview questions

IBM China Research Institute Intern

As early as March, I interviewed an intern at the IBM China Research Institute, mainly for natural language processing and semantic web. At that time, I was still attending school in Hong Kong. The two examiners interviewed me by phone for about half an hour. First, I introduced myself, then, I will introduce my learning experience and project experience, and then I will ask my questions based on these presentations, mainly about my algorithms and data structures, big data and data mining algorithms learned during the Master's degree, such as PageRank and LDA topic models, and clustering and classification, such as kmeans, DBSCAN, and decision tree, such as SVM, no questions are asked about neural networks. After all, I am an intern and have not asked too much, but the basics are still necessary. Then there is the classic algorithm, which is better than the algorithm. At that time, they asked me a question: a large-scale graph with millions of vertices and tens of millions of edges, how to efficiently find the shortest path between any two points.

SAP China Research Institute

This is social recruitment. I tried it at the time. Of course, it was also a telephone interview because I had an internship at IBM. The general process is also. I will first introduce myself first, and then learn and project experience. Next, the examiner asked me about the data structure and algorithm as well as the operating system compilation principles. First, how to write the non-recursive quick sorting and the non-recursive Binary Tree in-order traversal, then I asked me about the multi-level feedback queue scheduling algorithm, and then asked me about the compilation principle. c file and. I understood the differences between the H file as from the differences at the time of compilation, and later I found that the answer was wrong. Yes, from the perspective of the compilation principle.

Netease games

In this year's Netease game school, I chose business intelligence, which is actually big data analysis. I personally feel that the written test is relatively simple, like the Non-Recursive Implementation of Binary Trees, inputting any number, calculates the full arrangement from 1 to this number, which is non-recursive. There are also some pointer issues, such as linked list insert operations. The last question is a big system question, which is exactly the same as Baidu's pen question last year. It was designed as a distributed system.

On the other hand, I will introduce myself and talk about the project experience. At this time, the examiner asked me, Have you implemented all these algorithms in code? I said yes, I have implemented all these algorithms, at this time, he took a piece of paper and asked me to write out the main implementation code of the most basic kmeans algorithm. I did write this, So I soon wrote it out. Then the examiner asked me about my understanding of data analysis and asked me questions. If you are the boss in a busy area, how would you increase your profits, the focus is on a convenience store in busy areas, such as the Bund in Shanghai.

The second aspect is the final stage, where I asked questions on the site to solve the scheduling problem. If you are the boss of the express delivery company, within a 10 square kilometer area, there are 20 thousand packages to be sent or received every day, and one month is subject to the average distribution, find out how many couriers should be allocated to this region, then, design a reasonable scheduling model to efficiently implement express delivery scheduling in this region.

Later, I talked about my life and asked my hobbies, who I worship? What are my requirements for work cities? What are my salary expectations? When can I go to work.

IBM Xi'an CDL interview

After chatting for two and a half hours, from half past three to off-duty, I was also drunk. I asked both development and data analysis. In terms of development, I mainly asked about java and the design mode. In terms of java, I asked about the reflection mechanism, stream, java multithreading, the understanding of object-oriented, and the design mode. The examiner said, don't talk to me about the decorator mode and the factory mode. Then, ask me how to understand the design mode. Then, I have a question. If you want to implement a function, you can store it to a file, you can also store data in the database, or store data in other areas that will be extended in the future. How can this problem be designed. Data analysis is basically similar to the data mining aspects of the above companies, but I also asked more specific team Data Mining understandings and the main process of data mining, I also asked why I chose an algorithm in my project. For example, I asked why I chose kmeans for clustering instead of other algorithms. I also asked my understanding of big data, where is big data, and then how to improve e-commerce website product recommendations. Finally, I wrote an algorithm question, that is, input A number such as 3, and output A, B ,... z, AA, AB ,... ZZ, AAA, AAB, AAC ,... ZZZ, the input number is arbitrary, and then write the implementation code.

To sum up, in fact, the interview mainly depends on the direction of your position. The skill industry has a specialization, and the Foundation must be solid, so that it will not be so difficult to get an offer!

This article permanently updates the link address: