PS初稿求建议
20536
这个版块不知道可不可以发PS修改讨论的帖子，如果发错了地方请鹳狸猿删帖就好了哈~
我的PS感觉套路不太一样，请大家多多拍砖，多谢了！
[align="left"]Three years ago, when I watched PranavMistry’s talk about his SixthSense technology on TED in my first computerclass, I believed I have set my foot on the way of changing the world, but hadno idea with the specific method to reach the destination. After a vibration ofmy mind by a professor, along with the accumulation of experience and abilitiesthese years, I think I have found the entrance to make a difference. That isData Mining and Machine Learning.[/align]
I started to seek the path by joining aData Mining Group and commencing on a project of Music Recommendation, justbecause of a passion on coding and the enjoyment of applying theory to thereality. However I had a deeper contemplation after a machine learning class.The professor was introducing the background and development of artificialintelligence and machine learning. She mentioned that “There is a possibilitythat the model we have built and the method we have utilized in machine learningall the time are completely wrong. That may explain the imperfect of thetranslation system or the stagnant in many other areas. I hope that one daysomeone of you could propose a new system and make a change of the world.”
I was so shocked by the idea that I accessed more information about machinelearning immediately after class. The criterion of machine intelligence-Turingtest fascinated me mostly. It is miraculous and invigorating to imagine amachine communicating with a human with no obstacle. However, according to the hugegap between ideal and reality, maybe my professor’s view is reasonable and theresearch have deviated its direction. If I could deal with the problem from anovel angle, the real intelligence of a machine might be accomplished, and thechange of our world will beyond our imagination. So I further read manyarticles on Natural Language Processing, and felt that too much attention hasbeen paid on the analysis of particular sentences, which likes the process of analyzingmachine language and programming language. Affiliating emotional factors may bea breakthrough in NLP research.
I believe I have found a way to achieve my goal of changing the world by changingthe traditional research system first. I could hardly contain my excitement andwent to my supervisor for discussion. While he calmed me down and suggested meto do some fundamental works since scientific research is “not so effortless”.
[align="left"]To get a more definite understand of datamining and machine learning, my research focused on a basic skill - classification.I processed data of protein sequences in SCOP database, extracted 188dimensional features from each sequence and classified it into protein foldswith an ensemble classifier. I read most articles on protein fold recognition,from which I found most researches just combine several basic classifiers, unawareof their properties. Based on a previous study of my supervisor, the diversityof basic classifiers facilitates better classification result. Thus I utilizeda clustering algorithm and several selective strategies to choose a series of discrepantclassifiers. My ensemble classifier outperformed traditional classifiers, and Ifelt it fantastic and delightful to cluster classifier performances in aclassification problem. What is more important, I began to know that changescan be achieved in every step we are taking, instead of only in big issues.[/align] After getting over a chapter of unanticipated difficulties, a paper entitled “HierarchicalClassification for Protein Fold Recognition Using a Novel Ensemble Classifier”(which can be accessed in appended materials) was submitted to PLOS ONE onSeptember, 2012. I also developed a web server (which can be accessed at url)for application and discussion with more users. However, the most preciousacquisitions should be the improvement of my programming, reading and writingarticles abilities, and the deeper understand of how a scientific research islike.
During the research, I had developed bettercomprehension on my research area. When I search in Google, buy books onAmazon, and use Siri App in my iphone, I can always feel the changes in thereal life owning to data mining and machine learning. The research experienceand further understanding reshaped my goal to be no longer obscure orunrealistic. I still keep my enthusiasm in seeking innovations in machinelearning, but with a broader vision. I realized that every little change ofpeople’s life is a form of changing the world. So my final goal is obtaining anopportunity to communicate with latest ideas and enhancing my specialty levelto be able to change people’s lifestyle. I wish to continue my education in anadvanced environment of data mining and machine learning and I expect maybesomeday I indeed built a new system in this area and extensively changed theworld.
University of Illinois is famous for its studyon data mining and machine learning. Its research area covers most majorresearch directions, and stands out in many study categories. Students haveopportunities to secure an advisor and do research on their thesis, and thatattracts me extremely. Since I have a clear objective of my specialty and I amready for graduate study in both research and academic aspects – I rank top 20percent in 120 students and I have taken two extra mathematical courses tostrengthen my theoretical foundation – I believe I can exert my potential tothe most degree and have the maximum possibility to achieve my goal if I havethe chance to enter your university. I sincerely ask you to consider myapplication favorably.
我的PS感觉套路不太一样，请大家多多拍砖，多谢了！
[align="left"]Three years ago, when I watched PranavMistry’s talk about his SixthSense technology on TED in my first computerclass, I believed I have set my foot on the way of changing the world, but hadno idea with the specific method to reach the destination. After a vibration ofmy mind by a professor, along with the accumulation of experience and abilitiesthese years, I think I have found the entrance to make a difference. That isData Mining and Machine Learning.[/align]
I started to seek the path by joining aData Mining Group and commencing on a project of Music Recommendation, justbecause of a passion on coding and the enjoyment of applying theory to thereality. However I had a deeper contemplation after a machine learning class.The professor was introducing the background and development of artificialintelligence and machine learning. She mentioned that “There is a possibilitythat the model we have built and the method we have utilized in machine learningall the time are completely wrong. That may explain the imperfect of thetranslation system or the stagnant in many other areas. I hope that one daysomeone of you could propose a new system and make a change of the world.”
I was so shocked by the idea that I accessed more information about machinelearning immediately after class. The criterion of machine intelligence-Turingtest fascinated me mostly. It is miraculous and invigorating to imagine amachine communicating with a human with no obstacle. However, according to the hugegap between ideal and reality, maybe my professor’s view is reasonable and theresearch have deviated its direction. If I could deal with the problem from anovel angle, the real intelligence of a machine might be accomplished, and thechange of our world will beyond our imagination. So I further read manyarticles on Natural Language Processing, and felt that too much attention hasbeen paid on the analysis of particular sentences, which likes the process of analyzingmachine language and programming language. Affiliating emotional factors may bea breakthrough in NLP research.
I believe I have found a way to achieve my goal of changing the world by changingthe traditional research system first. I could hardly contain my excitement andwent to my supervisor for discussion. While he calmed me down and suggested meto do some fundamental works since scientific research is “not so effortless”.
[align="left"]To get a more definite understand of datamining and machine learning, my research focused on a basic skill - classification.I processed data of protein sequences in SCOP database, extracted 188dimensional features from each sequence and classified it into protein foldswith an ensemble classifier. I read most articles on protein fold recognition,from which I found most researches just combine several basic classifiers, unawareof their properties. Based on a previous study of my supervisor, the diversityof basic classifiers facilitates better classification result. Thus I utilizeda clustering algorithm and several selective strategies to choose a series of discrepantclassifiers. My ensemble classifier outperformed traditional classifiers, and Ifelt it fantastic and delightful to cluster classifier performances in aclassification problem. What is more important, I began to know that changescan be achieved in every step we are taking, instead of only in big issues.[/align] After getting over a chapter of unanticipated difficulties, a paper entitled “HierarchicalClassification for Protein Fold Recognition Using a Novel Ensemble Classifier”(which can be accessed in appended materials) was submitted to PLOS ONE onSeptember, 2012. I also developed a web server (which can be accessed at url)for application and discussion with more users. However, the most preciousacquisitions should be the improvement of my programming, reading and writingarticles abilities, and the deeper understand of how a scientific research islike.
During the research, I had developed bettercomprehension on my research area. When I search in Google, buy books onAmazon, and use Siri App in my iphone, I can always feel the changes in thereal life owning to data mining and machine learning. The research experienceand further understanding reshaped my goal to be no longer obscure orunrealistic. I still keep my enthusiasm in seeking innovations in machinelearning, but with a broader vision. I realized that every little change ofpeople’s life is a form of changing the world. So my final goal is obtaining anopportunity to communicate with latest ideas and enhancing my specialty levelto be able to change people’s lifestyle. I wish to continue my education in anadvanced environment of data mining and machine learning and I expect maybesomeday I indeed built a new system in this area and extensively changed theworld.
University of Illinois is famous for its studyon data mining and machine learning. Its research area covers most majorresearch directions, and stands out in many study categories. Students haveopportunities to secure an advisor and do research on their thesis, and thatattracts me extremely. Since I have a clear objective of my specialty and I amready for graduate study in both research and academic aspects – I rank top 20percent in 120 students and I have taken two extra mathematical courses tostrengthen my theoretical foundation – I believe I can exert my potential tothe most degree and have the maximum possibility to achieve my goal if I havethe chance to enter your university. I sincerely ask you to consider myapplication favorably.
- 3
6条回复