A routing strategy would result in a significantly higher

A tingle went down my spine as I read the banner, “Welcome to Robocon India 2013!”. I had spent the entirety of the previous year enabling Marvin for autonomous navigation in preparation for this moment. Standing in front of that banner with him and the rest of my team then, felt surreal. We were competing against the teams representing about ninety of the finest engineering schools in the country and I was excited to be there. In what was our first competition at this level, we ended up winning the award for The Most Cost Effective Design along with a place inside top ten. While I was elated with the achievement, this experience also served as an inspiration to learn further about embedding intelligence into machines and different ways in which it could be done.

 

This experience influenced my choice of elective subjects inside the classroom as well. I opted for Soft Computing Techniques in my final year of undergraduate studies, a subject in which I ranked among top 1% in the department. This subject introduced me to fuzzy logic and I couldn’t appreciate its power and simplicity more when I implemented fuzzy rule based speed control for Marvin’s autonomous navigation problem and observed that he performed as well as when the relatively complex PID algorithm was used earlier.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

 

The same subject played a pivotal role yet again when I learnt about neural networks and used the back propagation technique to build a character recognition model for English alphabet. As a part of another project, by considering the refractory effect of a chaotic neuron along with few other parameters, I learnt to build a routing strategy model with chaotic neurodynamics in communication networks. I realized that usage of this routing strategy would result in a significantly higher arrival rate of the packets compared to other routing strategies such as shortest path routing strategy. I consequently presented a paper on this project during a course as a part of my curriculum.

 

During my work tenure at Sonus Networks as Software Engineer, I was provided with ample opportunities to enhance and thereafter apply my knowledge at the system level and application level while working on various projects. As a part of the latest project, called SWe Configurator, I developed a core partition algorithm (for the most sought-after product of Sonus) from the scratch. By considering the complexities of the various parameters that influence the partitioning of cores, the algorithm allocates cores efficiently for all the half a dozen functions. My efforts in the creation of this optimal solution were recognized with an Individual Excellence Award. To add to my delight, I was informed that I was the youngest employee to win this award.

 

Outside of my professional life, my earlier experiences with neural networks motivated me to further explore the vast field of machine learning. After attending a MOOC on Machine Learning, I applied the newly attained knowledge while working on various side projects. In one, I used SVMs to build a spam classification filter. Furthermore, I developed a personal movie recommender application by making use of collaborative filtering learning algorithm. Working on these projects gave me a great insight into the power of machine learning and how it could be applied to diverse fields.

 

As happy as I was with developing optimum solutions for the communication network problems at Sonus Networks, over the last couple of months, I realized that solving problems while applying machine learning would be more fulfilling personally. However, on a couple of occasions, I was confronted with difficulties while preparing the data sets. In addition, I was intrigued by how one can turn the developed model into a data product. I recognized that to gain further insights into machine learning and how I could effectively use it to solve the modern-day problems (in the fields of robotics and networks) in the industry, I needed to get back to school.

 

The remarkable research work carried out in the field of machine learning at the university played a major role when I finalized to apply for University of Florida. It would be a great privilege if I get a chance to contribute to research at the university as my interests lie in his field. Inside the classroom, courses such as Advanced Machine Learning & Analysis of Algorithms would provide a comprehensive knowledge and thereby help to strengthen my fundamentals.

 

I believe that my strong mathematical background (a subject in which I stood among top 8% in my department) and the knowledge that I had gained while attending a MOOC on Algorithms & Data Structures would play a crucial role in helping me succeed at the graduate school. I feel that under the able tutelage and guidance at the university, I would realize my aspirations. Given the opportunity to be a part of your promising graduate study, I shall bring credit to the institution.