As a researcher, my research areas and interest lies towards AI, machine learning techniques like Regression, Classification, Deep learning, CNN, RNN and compilers, operating systems, computer architecture. Specially on optimizations techniques and acceleration of various kinds of applications, like CNN, Bio-metrics etc.
I have invented a new instruction selection technique for compilers, reconfiguring CNN layers for acceleration and reduction in hardware for custom implementation of deep learning on silicon and currently working on a unique convolution algorithm, with patent pending.
Theory is important for making foundations, but what more important for me is to learn from practice, i.e. implementing the ideas, evaluate them and modify them and do it again, unless I have achieved what I was set out for.
o Natural Language Processing.
o Text Summarization.
o Sentiment Analysis.
o Generative Adversarial Networks.
o Meta Learning.
o Recurrent Neural Networks.
o Convolutional Neural Networks.
o Regression and Classification.
o Operating systems.
o Acceleration and Optimization.
P.S. Check out my website at https://vijaydaultani.github.io/