Overseas Internship Scheme (OIS)

"My supervisor gave me guidance on how to express an idea in a more structured way. This process of learning enhanced my academic writing skills. During my research, another important thing is results. Even though we provide many mathematical proofs in our report, we still need to have a good result to support our framework. I spent time adjusting the parameters to have a better result. This experience gave me a better understanding of how to design and test an experiment in machine learning. Since we need many hours for training the data, we need to adjust many details of the implantation. For example, how to read the data? Which method is more computational efficient? During these details of the designing an algorithm. It provided useful skills for designing a better algorithm in terms of computation time."

"This outreach activity provided me the opportunity to have a US research experience, which helped me discover my interest in image processing. I studied Principal Component Analysis (PCA), t-SNE (t-distributed stochastic neighbor embedding), and CNN (Convolutional Neural Network) using Python. I learned to look for trained models online and modify them instead of building up from zero, which often saves time and has good performance. Additionally, my skills in writing English reports, communicating in English, and preparing for presentations have improved in constant practice. Being in an English-speaking environment motivates me to make these improvements. The questions my peers asked after my presentation made me think more about my methods. I learnt that I should comprehensively understand my project before discussing it with others. I also realized the importance of a background introduction, which is the premise of whether others can make sense of the follow-up contents."

"Platforms make people, and people success people. In the daily cooperation and exchanges with Prof. Ke Chen at University of Liverpool, I have gained a lot of new perspectives. This is different from traditional "knowledge" on the textbook, the same knowledge may get completely new results from a different perspective. And this, in turn, helps us better understand the whole system and points to where the path will go in the future. That's the beauty of communication in research. Through the opinions and ideas from our peers, supervisor, and collaborators, we can create more interesting things and understand what we have mastered more comprehensively. These are usually hard to come by, which is exactly what the fantastic platform could provide us: glorious ideas, great mentors and peers."

"One of the best part about this internship was the research group I was working with. I got to work with and learn from Professor Nikolas Provatas and his group who are on the leading edge of phase field models (a powerful computational model in physics and material science); as a matter of fact Professor Provatas is one of the leading authorities on this topic and he wrote one of the most comprehensive books on it. The group members were also some of the most talented and hard-working people I have met, each working on a different project that is on the frontier of phase field modelling. Each of their project would have significant impact on the topic."

"The programme is very well-structured. We made two presentations and attended a poster symposium. The two presentations provide us a terrific opportunity to get familiar with what our peers have been working on and get some insights from other people’s talk. Apart from mathematics projects in our programme, there were also some other projects displayed in the poster session. Presenting my work and listening to other people’s work helped me to figure out the problems we had.

In addition, I also travelled to several places with my peers during holidays. I went to Orlando on the Independence Day and New York City after the programme ended. I’m crazy about classical music. I was quite excited that I had a chance to go Carnegie Hall to listen to Bach’s Cello Solo Suites in live. It’s been a dream since when I was a little kid. The performer talked about composition of the suites and the story about the cello he played. Besides, I also went to the Metropolitan Museum of Art where I saw a real Stradivari violin in person.

In conclusion, the wonderful summertime in the U.S. is definitely a memorable experience. Not only did I have some research experience, but also explored some new places."

"Besides all the techniques and trainings in cancer research, since this is my first internship, I have also learnt a lot on how to communicate with my colleagues and supervisors, which before going to Australia I always have a difficult with, as I always get so nervous whenever I am facing my professors. I also think that the internship provides me with a much better idea of what it is like to work full time in a lab during research, and this helps me to plan for my future path of whether I actually want to spend the rest of life during research.

All in all, I have grown a lot and learnt a lot throughout the whole internship and I believe all this new knowledge and experiences will be a nice base for my future studies and even my future career to build on."

"I experienced a short time of researcher life. My thought and behavior changed a lot. Before this trip, I think that concentration is the most important thing for a researcher. However, this experience told me that communication is really significant in research. The most obvious evidence is that when we try to express our ideas, we attempt to make our thoughts clear and logical because we need to explain it to other people and we find new ideas in existing things. Even though we just repeat the sentence we wrote, there are still new things hidden in unimposing corners.

In conclusion, this experience broadened my view. My consideration towards my future researcher life becomes more comprehensive. Besides that, I have a more exhaustive cognition of myself."

"During this summer, I first reviewed the literature about convex geodesic image segmentation, then tried to convert the MATLAN code of the numerical algorithm into Python version. Finally, I implemented some supervised deep learning framework and proposed an unsupervised deep learning model. The content about developing a deep learning framework, which aims to have similar effect as the numerical algorithm, is very popular in these years. But sometimes it is hard to implement and explain the meaning of the framework because of the lack of mathematical explanation of the deep learning. Therefore, I think mathematics is quite important in nearly all fields of study. However, the deduction of mathematical formula sometimes is boring and needs quite a lot training. I think I lack this ability and do not have sufficient mathematical background, though I major in Computing Mathematics.

Besides this, I find that I have much more interest in the practical applications to solve some real problems instead of the deduction or create some mathematical formula. So I may keep studying and turn to some practical areas like Computer Science or other engineering areas. I am very grateful for this opportunity and the chance to experience the England study life."

"This summer I joined a study in using machine learning to deal with a problem in sparse matrix. Sparse matrices often occur in real problems. However, when we are solving a linear system with sparse matrix as coefficient, the matrices may become less sparse during the process of LU decomposition. The added non-zeros in the LU decomposition are called fills. If we could find a method to reduce the fills, we may save a lot of memory in actual computing.

In practice, there are some methods trying to solve the problem. Among them there is one called nested dissection. This method considers the sparse matrices as graphs and aims to find a sequence of separators and each separate the graph into two parts so that there won’t be fills occur between the separated parts. By finding these separators, we hope the fills can be reduced.

Looking back, this internship was helpful. Through this internship, I learned a lot of new knowledge in machine learning and sparse matrix. Also this experience in Berkeley lab help me know more about the method to deal with a mathematical study in a group. My communication skill is also improved through this summer and I met friends from various cultures."