MS PS求改，马上要提交，已改几遍，求各位大大帮忙再看看，略虚。申EE和CS
23405
Personal Statement
(CS)I love computer games, but I am never addicted due to my strong control. This hobby leads me to the world of computers originally, and makes me fascinated with computers’ capacity of constructing scenes. Also, when I saw the RoboCup I was infatuated with the intelligence of robots. I believe computers can do anything and hope to produce something useful to our society. Now being well equipped with keen interest and solid academic ground, I am writing to apply for admission to the Master of Science in AAAA program at 某学校某系.
(EE)I love movies. When I was young I thought they just simply contained a series of images. After my high school study, I know they are pretty complex, which contain intraframe compression and interframe compression and other technology. If they simply contain many primitive images, their size will be much bigger, which would limit the development of movies greatly. This leads me to the fascinating world of signal processing, for it can change our world. Now being well equipped with keen interest and solid academic ground, I am writing to apply for admission to the Master of Science in AAAA program at 某学校某系.
When I was a freshman in my university, I was interested in all kinds of curriculums especially Calculus. Due to my solid background of math, I got high scores both in Single Variable Calculus and Multivariable Calculus. Besides, I got second prize at National Math Contest for College Students in Anhui Province.
In my third year, I learned Digital Signal Processing (DSP). The Professor Guo Wei required us to accomplish a project with a group of 4 and I became a leader of a team. At that time two members of my team were going to participate TOEFL, so as the leader I needed to carry more duty. I accomplished the most algorithm to construct filters and analyze audio. In addition, I built a system to achieve some functions like analyzing dial tone and giving analysis of audio automatically. I directed a discussion to review what we have finished and decide what to be achieved per week which ensured all of members could make progress together, and eventually they also got an A for this curriculum.
To broad my view, I entered Multimedia Computing and Communicating Lab in the third year and made some professional research following with my mentor, Weiming Zhang. He assigned me steganography on image and video. Strongly encouraged by my intense curiousness and expectation of my mentor, I worked hard and quickly learnt how to hide bit stream with a simple definition of distortion. Then I created my own distortion function. To evaluate it, I trained a classifier with the knowledge of machine learning and got the detecting error rate which was expected to be high. In further study I also learnt some new definition methods like Non-local and used them in my work.
To promote my research skill, I participated in Undergraduate Research Program of USTC that defining the distortion by predicting the quantizing direction of DCT coefficients in JPEG image. I proposed a method of making DCT coefficients plus and minus one respectively have a DCT inverse transformation and comparing the variance of spatial matrix. Via many coding experiments I found on many images the right rate of prediction of “one” can be nearly to 93%. The huger the difference of those two variance is, the higher the right rate of prediction is. But that is not enough because “zero” is the most elements in DCT coefficients of JPEG images and the right rate of prediction of “zero” is only nearly 63%. Not satisfied with these experiment consequences, now I am still working on this topic and hope to find other methods to promote the accuracy of predicting “zero”.
In summer of 2013, I learnt Statistical Data Modeling taught by Professor Hui Jiang of York University concerning about machine learning and pattern verification. I was very interested in machine learning. I did two projects. Firstly, I built a classifier based on Gaussian Model and optimized it with minimum classification error (MCE), then measured the performance with test set. Secondly, I built a bi-gram word language model based on the training text data with Katz’s back-off scheme, then evaluated the performance by calculating the perplexity on a set of test data. When I accomplished my projects, I did not regard them as the final results. After asking the professor and our teaching assistant for reference value and modifying my program for many times, I eventually got my conclusion and an A for this class. In my opinion, doing research requires courage to handle every unexpected problem with great patience, diligence and intelligence.
Considering my interests and capabilities, I believe the AAAA program at AAAA University will be an ideal place for me to pursue a master degree, because the research program focuses on AAAA, which matches my background well. In addition, I am attracted to the emphasis on AAAA and the core curriculum like AAAA. As I am determined not only to conduct scientific research but also to apply advanced scientific findings in real-world cases, these abilities are crucial for my professional success. After the graduate studies, I want to pursue a doctor degree. I believe I will best capitalize my research acumen and past experience at 学校 and successfully achieve to be a world-class researcher.
(CS)I love computer games, but I am never addicted due to my strong control. This hobby leads me to the world of computers originally, and makes me fascinated with computers’ capacity of constructing scenes. Also, when I saw the RoboCup I was infatuated with the intelligence of robots. I believe computers can do anything and hope to produce something useful to our society. Now being well equipped with keen interest and solid academic ground, I am writing to apply for admission to the Master of Science in AAAA program at 某学校某系.
(EE)I love movies. When I was young I thought they just simply contained a series of images. After my high school study, I know they are pretty complex, which contain intraframe compression and interframe compression and other technology. If they simply contain many primitive images, their size will be much bigger, which would limit the development of movies greatly. This leads me to the fascinating world of signal processing, for it can change our world. Now being well equipped with keen interest and solid academic ground, I am writing to apply for admission to the Master of Science in AAAA program at 某学校某系.
When I was a freshman in my university, I was interested in all kinds of curriculums especially Calculus. Due to my solid background of math, I got high scores both in Single Variable Calculus and Multivariable Calculus. Besides, I got second prize at National Math Contest for College Students in Anhui Province.
In my third year, I learned Digital Signal Processing (DSP). The Professor Guo Wei required us to accomplish a project with a group of 4 and I became a leader of a team. At that time two members of my team were going to participate TOEFL, so as the leader I needed to carry more duty. I accomplished the most algorithm to construct filters and analyze audio. In addition, I built a system to achieve some functions like analyzing dial tone and giving analysis of audio automatically. I directed a discussion to review what we have finished and decide what to be achieved per week which ensured all of members could make progress together, and eventually they also got an A for this curriculum.
To broad my view, I entered Multimedia Computing and Communicating Lab in the third year and made some professional research following with my mentor, Weiming Zhang. He assigned me steganography on image and video. Strongly encouraged by my intense curiousness and expectation of my mentor, I worked hard and quickly learnt how to hide bit stream with a simple definition of distortion. Then I created my own distortion function. To evaluate it, I trained a classifier with the knowledge of machine learning and got the detecting error rate which was expected to be high. In further study I also learnt some new definition methods like Non-local and used them in my work.
To promote my research skill, I participated in Undergraduate Research Program of USTC that defining the distortion by predicting the quantizing direction of DCT coefficients in JPEG image. I proposed a method of making DCT coefficients plus and minus one respectively have a DCT inverse transformation and comparing the variance of spatial matrix. Via many coding experiments I found on many images the right rate of prediction of “one” can be nearly to 93%. The huger the difference of those two variance is, the higher the right rate of prediction is. But that is not enough because “zero” is the most elements in DCT coefficients of JPEG images and the right rate of prediction of “zero” is only nearly 63%. Not satisfied with these experiment consequences, now I am still working on this topic and hope to find other methods to promote the accuracy of predicting “zero”.
In summer of 2013, I learnt Statistical Data Modeling taught by Professor Hui Jiang of York University concerning about machine learning and pattern verification. I was very interested in machine learning. I did two projects. Firstly, I built a classifier based on Gaussian Model and optimized it with minimum classification error (MCE), then measured the performance with test set. Secondly, I built a bi-gram word language model based on the training text data with Katz’s back-off scheme, then evaluated the performance by calculating the perplexity on a set of test data. When I accomplished my projects, I did not regard them as the final results. After asking the professor and our teaching assistant for reference value and modifying my program for many times, I eventually got my conclusion and an A for this class. In my opinion, doing research requires courage to handle every unexpected problem with great patience, diligence and intelligence.
Considering my interests and capabilities, I believe the AAAA program at AAAA University will be an ideal place for me to pursue a master degree, because the research program focuses on AAAA, which matches my background well. In addition, I am attracted to the emphasis on AAAA and the core curriculum like AAAA. As I am determined not only to conduct scientific research but also to apply advanced scientific findings in real-world cases, these abilities are crucial for my professional success. After the graduate studies, I want to pursue a doctor degree. I believe I will best capitalize my research acumen and past experience at 学校 and successfully achieve to be a world-class researcher.
- 3
5条回复