Statement of Purpose to pursue graduate studies at UT Austin in Computer Science Engineering.
STATEMENT OF PURPOSE
As a kid, I was really inspired by the explorers and I wanted to be an explorer when I grew up. As a computer engineer, I would always look for new things that we can do, that weren't just ever possible. Machine Learning feels like an intellectual exploration. With the advent of lot more data and lot more computing power we can really do things that weren't envisioned before. The uptake of Machine Learning in the last five years opened my eyes to the immense potential of Machine Learning which could give machines innate abilities to learn, just as humans learn. This shed insight on my thinking and made my thinking more enjoyable. I enjoyed the experience immensely and it was quite successful in driving me towards Data Science and I am in pursuit of advanced level knowledge in this field. I really hope to get enrolled into the Master of Science in Computer Science (MSCS) program at The University of Texas at Austin. I have summarized my educational background and my motivation for choosing UT Austin for graduate studies in the following paragraphs.
My interest in Mathematics goes back to the time when I was at school. The interest has only grown through my years in school and college, as I have learnt more and more about the subject. My exposure to Computer Science began after I joined the Indian Institute of Information Technology Design & Manufacturing (IIITDM) Kancheepuram, to pursue my under graduation. The unique factor that attracted me towards this college is its interdisciplinary curriculum which gives a hard core computer science student like me enough insight into other branches, which made me appreciate the importance of computer science in every field.
My undergraduate career started with an acute interest in Application Development along with my predilection towards Coding. I was involved in the Web Operations team at IIITDM, which develops and maintains many web applications running in the institute. While being a part of the Web Operations team, I developed several applications including the web portal for the institute's techno cultural fest, online faculty and staff recruitment portals with automated process of shortlisting applicants, an online student portal called 4pi for the students to share information, organize events, conduct polls and create academic profiles that can be shared with anyone across the globe. The summer of second year provided me with an opportunity to solve the problem of inaccessibility of Devanagari Fonts. I worked on designing and developing a web based font tool to improve the accessibility by creating a data hub consisting of several Devanagari Fonts under the guidance of Dr. Girish Dalvi of the Indian Institute of Technology Bombay (IIT Bombay). This tool, an interactive visual-feature based font search system, provides an intuitive and user-friendly method to search and browse through a large number of fonts.
Having been exposed to the various facets of Computer Science in the course of my undergraduate studies, I have found Machine Learning most intellectually satisfying and simulating. Courses like Data Mining and Machine Learning have been in my firm favourites. I find Machine Learning particularly appealing because of the flavour of Linear Algebra and Optimization that I love so much. At the same time, the application of Machine Learning techniques to solve practical problems require exploration and research. This adds to their appeal. With the advancements in the areas of Deep Learning and Artificial Intelligence, the applications of this field in various domains became more interesting and challenging. Members of the faculty at The University of Texas at Austin have done a lot of pioneering work in these areas.
During my elective course on Data Mining, I explored the applications of Data Mining techniques in the image compression domain. I worked on Lossy Image Compression and developed a Data Mining based compression algorithm which clusters similar pixels in the image and then uses the cluster identifiers to mine closed frequent sequences. The encoding process is optimized by refining the conventional Generalized Sequential Pattern Mining(GSP) algorithm, to achieve significant reduction in the compressed image size. This algorithm for image compression was published as Avinash Kadimisetty, C. Oswald and Sivaselvan B., "Frequent Pattern Mining Approach to Image compression", in the proceedings of 22nd IEEE international conference on Advance Computing and Communication (ADCOM), Bengaluru, India, September 2016.
As a part of my undergraduate major project, I extended the research in Lossy Image Compression by employing efficient clustering and parallelising the encoding process to improve the efficiency in time and space. Simulations of the proposed algorithm indicate significant gains in compression ratio and quality in relation to JPEG and GIF. This algorithm has achieved better compression than GIF algorithm, decreasing the compressed image size by ~18%. This research work was published as Avinash Kadimisetty, C. Oswald and Sivaselvan B., "Lossy Image Compression - A Frequent Sequence Mining perspective employing efficient Clustering", in the proceedings of 13th India Council International Conference (INDICON), Bengaluru, India, December 2016.
After graduating from IIITDM, I chose to work in the field of Machine Learning to learn to handle big data fluently and get hands on experience on real world projects. I am currently working as a Junior Data Scientist at Evive Health for close to a year in the field of Machine Learning and Artificial Intelligence. At Evive Health, I was involved in the development of several projects to monitor a person's health activities and target interventions using ML and AI. A project I would like to highlight is Hospital Readmission, which aims to reduce the number of unplanned readmissions, thereby reducing the financial burden on patients. Most of the readmissions occur due to improper care and irregular usage of medicines after discharge. As a part of this project, I built models using Machine Learning techniques like Logistic Regression and Random Forests to find out the probability of a person getting readmitted once discharged. The patients at high risk of readmission are notified and are advised to consult a doctor to reduce the chances of readmission. Alongside these projects, I am exploring the applications of Deep Learning in the healthcare domain.
For me, problem solving is an exhilarating experience, unequalled by anything else. It would be immensely gratifying for me to able to contribute something to the understanding of a subject that has given me so much pleasure and joy. I look forward to a career in research where not only my academic background will be used to achieve my research goals but also make original contributions to my field of interest. My undergraduate projects and my work at Evive Health have played a major role in my decision to pursue higher education in the fields of Machine Learning and Artificial Intelligence. Having done my undergraduate studies in undoubtedly one of the best undergraduate institutions of India, I would consider it my privilege to be able to pursue my graduate studies at The University of Texas at Austin and avail the excellent infrastructure facilities and research opportunities it has to offer. The University of Texas at Austin suits my research interests because of its advanced academic curriculum which provides breadth of knowledge as well as command of a specific area of interest emphasizing on research methodology. I am sure that the stimulating academic environment and interaction with the distinguished faculty will prove immensely fruitful and facilitate my growth as an individual researcher in my field of interest.
Finally, I would like to thank the admissions committee for taking the time to review my application.