Edwin Kay

Background

About Edwin Kay

More than 10 years of research in physics and astronomy, with years of experience in astronomical data analysis and scientific inferences. Highly experience in software development, database management and designing graphical interfaces in multiple platforms. Actively learning and implementing artificial intelligence and data analysis/visualization tools. Quick learner, hard working and adaptive to different working environments.[You may find my publications and many of my professional contributions under the name "Ehsan Kourkchi."] I pursued my childhood dream to learn about the Universe and therefore I mostly acquired my knowledge and skills through my researches in the field of Astronomy. My researches benefited a lot from the novel data science and machine learning methodologies. I am excited to leverage my expertise and skill and offer new perspective in solving problems in industry. I am always open to explore new ideas and challenges. I am proficient in Python, C/C++, PHP and Java, JavaScript, IDL, HTML, Linux, SQL, Data Analysis, Numerical Simulations, Visualization, Statistics, Data Structure and Algorithm, Shell Scripting, GitHub, OpenMP & MPI. I am familiar with Machine Learning techniques and tools including Regression and Classification methods, Gaussian Process, Bayesian Inference, Random Forest, Deep Learning, TensorFlow, Scikit-learn, Time Series Analysis, etc- I have implemented, trained, tested and deployed a simple multi-layer convolutional neural network using the TensorFlow Python package to evaluate the 3D spatial inclinations of spiral galaxies- I developed a web application that provides the relationships between the distances and velocities of galaxies - I developed an online GUI,“Galaxy Inclination Zoo", that enables citizen scientists to participate in measuring the inclinations of spiral galaxies by evaluating their images - I use Bayesian statistical inference to explore the probability distribution of model parameters given the observational data - I have constructed non-parametric models based on the Gaussian Process methodology, to model the global dust obscuration in spiral galaxies as a function of their observable features - I have performed Principal Component Analyses to combine a set of mutually correlated galaxy features to evaluate their properties - I applied Chi-squared minimization and regression methods for model fitting - I used two-point correlation statistics to evaluate galaxy clustering and Fourier transformations for noise reduction