Unanswered [2] | Urgent [0]
  

Home / Letters   % width Posts: 4

Motivation letter for Master's program in Data Science (no specific instructions for the content)


Diahr 1 / 3 2  
Jan 4, 2022   #1
Good day, everyone.

I have written a motivation letter but I am not sure about the contents. In the letter, I tried to present a story as to how I came to the desire to apply for the program. However, I feel that it may be too focused on the progress rather than on the results. Yet, I am not sure if it would be better to elaborate further on past achievements since they are already mentioned in the CV and several letters of recommendation. For instance, in the 4th paragraph, I only mentioned that I overcame a problem when in fact, it resulted in a publication, which got the best paper award.

I will be glad for any feedback. Thank you in advance for your work.

The directions from the university are as follows:

The motivation letter should be written in English. The maximum length of the motivation letter is about 4000 characters. There are no specific instructions for the content of the motivation letter except for the general advice that you should write a separate motivation letter for each study option you apply to.

THE MOTIVATION LETTER:



Dear admissions committee,

During my second year at the university, I had an assignment for an English class: to present one's research essay before classmates. However, no one had slides at hand due to sudden schedule changes, so we decided to deliver it the old-fashioned way---with blackboard and chalk. The presentation, as I lectured, drew figures, interacted with the audience, quickly became a memorable experience. It made me realize that I wish to conduct research and share it with others.

Ever since I have been looking for opportunities to participate in academic events. One week, our department had a visit of [visiting professor] from [university A] to present his non-profit educational project [project A] to students of [city]. I loved the project's idea of teaching children from all around the globe to code, so, having asked the professor from our side, I was appointed as [visiting professor]'s assistant. The chance provided me with unique experiences: accompanying [visiting professor] around the campus; helping him prepare presentations; as an interpreter, assisting communicate with the audience; further, volunteering as a software developer. All this sparked my interest in international collaborations, strengthening the desire to become a professor one day.

Because of the new interest, I started to look for opportunities to spend a semester abroad. My application for the [scholarship name] scholarship was successful, and I began an exchange at [university B]. There, to my content, I was able to take master's courses in computer vision and statistical modeling, getting precious knowledge in areas related to machine learning, which has always interested me. Over time, I had become confident enough in the material of both courses and wanted to apply it in the real world. Therefore, having a goal of conducting research and, if possible, strengthening collaboration between universities, I came to ask a practice sessions assistant about the possibility of volunteering in the computer vision laboratory. Several days after, I was appointed to the [project B] project to save endangered [A and B species] ringed seals.

Since then, I have been working on ringed seals pattern patch matching: participating in weekly meetings, helping to write articles, independently investigating other re-identification approaches. The investigation led me to global pooling techniques used in image retrieval, which substantially improved the patch matching accuracy. However, they have the drawback of losing information about the spatial distribution of features. To overcome it, I employed knowledge gained from the statistical modeling course. Presently, I wish to further research representation learning by combining computer vision and statistics. Specifically, I want to investigate the applicability of Bayesian methods to CNN features to ensure robust metric and continual learning.

I believe that [target program]'s doctoral track is the best place for me to contribute to the field and to strengthen international cooperation. First, there is a computer vision professor [professor A] whose doctoral students have been working on problems closely related to that of my interest: semantic matching, image retrieval, continual learning. Therefore, I think I can be a valuable asset to his research group, further investigating these topics from a slightly different angle. Second, [target university] has a Probabilistic Machine Learning group with world-class researchers in Bayesian Deep Learning. Considering [target university]'s focus on collaboration, I am confident that we could work together on such interdisciplinary research, sparking ideas in each other. Lastly, [target university]'s emphasis on internationalization would give me an excellent opportunity to expand my international research network. Thus, I could give back to the world by increasing the number of research and educational collaborations around the globe.

Thank you for your time and consideration. I am looking forward to us pushing frontiers of computer vision together.

Yours faithfully,

[Name]
Holt  Educational Consultant - / 12,303 3989  
Jan 4, 2022   #2
Well, the motivational letter should reflect the reasons for the next step in your career path. It should not be a personal statement where you tell your academic biography and successes. Rather, your current career path should be dictating the content of the motivational letter stemming from the next step in your career (position promotion), solving a problem related to the field you work in (professional motivation), and a lack of theoretical knowledge to help you better perform your task (academic relevance in relation to work needs). These are the simple factors that the reviewers will consider when reading the letter of motivation. As of now, your work is more biographical in content, lacking in acceptable motivating factors. Simply being a person who loves / is interested in learning or continued education does not make for a convincing motivation for academic acceptance.
OP Diahr 1 / 3 2  
Jan 7, 2022   #3
Good day, everyone.

I once again post, hoping for your advice. In this letter, I tried to mention professional reasons for applying to the program. However, I am not sure about the introduction paragraph. Should I restructure it?

Thank you in advance.

MOTIVATION LETTER

Dear admissions committee,

With this application and my background in computer vision, I would like to get an opportunity to study in [program] to get a firm theoretical foundation in statistics. I believe that it will help me to achieve my goal: becoming a professional researcher in computer vision to work on continual unsupervised learning.

Two years ago, I studied at [University A] with the support of the [grant] scholarship. Throughout, I got experience in master courses related to machine learning: computer vision, statistical modeling, and GPU programming. Having a great passion for conducting research in the related fields in the future, I secured the highest score for all of the courses. Moreover, I got a volunteer position at the computer vision laboratory to work on the [project] project under professor [professor A]. The project, on which I have been working ever since, aims to preserve endangered [species] ringed seals through automatic photo-identification.

In [project], we have an abundant amount of raw data generated each week, with only a handful being labeled. In particular, the project relies on images taken from camera traps, which shoot seals from various angles. These images always contain different fragments of fur pattern, so our algorithm depends on the patch matching step: finding similarities between small parts of the pattern. This step calls for a large, labeled dataset, which is quite cumbersome to obtain in the case of small patches. After months of work, I understood how crucial such a dataset can be and how notoriously costly its creation is. For this reason, I wish to find a solution. I strive to make use of raw image data in such continuous systems.

I believe that the work on the problem will benefit various real-world applications with continuous access to new data. For instance, biologists may find it valuable for animal monitoring as the interest to go from physical tagging to non-invasive camera traps is only increasing. Consequently, the research could save endangered subspecies of giraffes, tigers, ringed seals, and more. Furthermore, different machinery, such as self-driving cars, drones, industrial robots, may find it useful for navigation. For example, presently, vehicles mostly rely on LiDARs, or laser-based radars, which are much more expensive than cameras.

I suppose that to solve the problem one should investigate continual unsupervised learning. This, however, poses the all-known problem of catastrophic forgetting---a network forgetting previously learned distribution upon meeting new samples. My work in [project] has proven that a fresh perspective from a statistical point of view may be a valuable source of inspiration for computer vision. Case in point, I utilized knowledge from the statistical modeling course to overcome the drawback of global pooling layers---a loss of information about the spatial distribution of features. The subsequent publication of [method] secured the best paper - 1st runner-up award. Therefore, I suppose it is possible to tackle catastrophic forgetting by applying Bayesian methods to CNN features. I hypothesize that one can learn the distribution along with feature mapping during the training process through MCMC algorithms.

I believe that [program]'s doctoral track is the best place for me to become a researcher and contribute to the field. First, there is a computer vision professor [professor B] whose doctoral students have been working on problems closely related to that of my interest: semantic matching, image retrieval, continual learning. Therefore, I think I can be a valuable asset to his research group, further investigating these topics from a slightly different angle. Second, [target university] has a Probabilistic Machine Learning group with world-class researchers in Bayesian Deep Learning. Considering [target university]'s focus on collaboration, I am confident that we could work together on such interdisciplinary research, sparking ideas in each other.

Thank you for your time and consideration. I am looking forward to us pushing frontiers of computer vision together.
Holt  Educational Consultant - / 12,303 3989  
Jan 8, 2022   #4
You have deviated totally from a motivational letter and created a statement of purpose instead. You cannot use this as a motivation letter but, you can set this aside for future editing when a statement of purpose is needed. Like I said, a motivational letter summarizes your academic, professional, and personal growth interests based on your desire for higher academic studies. Explain why you have chosen to go back to academic interests at this point in your career, in your life. All of which can be completed in a simple letter of no more than 5 paragraphs. Just follow the content discussion I am instructing you to use so that you can develop a general motivation letter of applicable information.


Home / Letters / Motivation letter for Master's program in Data Science (no specific instructions for the content)