SUBRAT KUMAR SWAIN

I have dedicated most of my time in my BTech career towards machine learning. In this 3 year, I have developed my appetite for deep learning. And I am very good at it.
The thing that drives me towards deep learning is the mathematical background of deep learning. How the optimization techniques are developed and how a simple tweak in the hyperparameter affects the end result and accuracy. Apart from that, I am also fascinated by its application in every phase of life ranging from speech, image to text audio and videos. Also recent advancement in this field like Deep RL, Meta-learning really drives me to work in this field of science and gain more and more knowledge about it.
Apart from using libraries, I have coded many of the deep learning architectures from scratch using no libraries. Some of them are restricted Boltzmann machine and deep belief network. I am currently working on a project named “Mining Socio-economic Factors Affecting Agricultural Productivity in Sambalpur District, Odisha State: Soft Computing based Machine Learning Approaches” funded by Science and Engineering Research Board, Department of Science and Technology, Government of India. I have also done research on the classification of continuous data using backpropagation based deep belief network to apply on the above project. I am also going to publish a journal on this research in between March and April. Along with that, I am also going to publish a survey paper on the same. This is the most complicated research project I have done so far.
From this project, I have learned quite a lot. Some of them are, how to apply deep learning model to a dataset, how to write efficient code from scratch without the use of any deep learning library, which helped me to know the nitty-gritty details of the architecture with minimal abstraction. Most importantly I have learned how to study the behavior of the dataset for a particular model and make changes accordingly to minimize the error and maximize the accuracy.
I know how to choose appropriate models for real-life problems from my experience of solving real-life problems during my research. I have also done projects on web development(Bookshop Automation System). So I can also represent my work through graphical interfaces like web and mobile and make it easier for all range of user to use it through the user-friendly graphical user interface. This will also help me to integrate systems and form an ecosystem that will work collaboratively to fulfill a common goal.