Hi, my name is Armin Foroughi, I'm a machine learning researcher in Biochemistry.

• University of California, San Diego San Diego, CA
  Bachelors of Science in Biochemistry with specialization in chemistry / Minor in Data Science
• Relevant Coursework: computational concepts (recursion, OOP), data structures (arrays, linked lists, stacks, queues, priority queues, heaps, binary trees, and hash tables), machine learning (Bayesian Decision Theory, SVMs, linear and Nonlinear Optimization, Deep Learning (CNNs, RNNs)), graph theory, probability, abstract data types, interfaces, algorithms and complexity, Spectrometry(NMR, IR, Mass Spec), Quantum Mechanics.

• Languages: Python, Java, SQL, R, C++, Bash, Git, Octave
• Cloud Computing: Amazon Web Services
• Computational & Visualization Libraries: Numpy, Pandas, Sklearn, TensorFlow, PyTorch, Scipy, Seaborn, OpenCV

Machine learning researcher

• Responsibilities: I perform a variety of data analysis related tasks, such as making graphs, image processing and building machine learning algorithms. I mainly work with .Star and .MRC format files developed from a Cryogenic electron microscopy and run my algorithm in the schools supercomputer through ssh cloud computing.
• Writing algorithm for MRC files: MRCs are a grayscale picture representation of a single protein. I wrote an algorithm that reads the MRCs as a 2D array then it rotates and translates each MRC and uses in place comparison to another MRC to find its closest MRC match. This creates an ordered single line of all MRCs that will be used to construct a 3D model of the protein.
• Writing algorithm for Star files: Star files are data frames of particles and their features that go into each MRC. I’m working on an algorithm that uses a three layer neural network to identify the best particles that go in each MRC and the bad particles that should not be used. A good particle is when it stays with one MRC and a bad one alters between MRCs. The algorithm looks at each iteration that CryoEm runs to form an MRC, and uses sigmoid activation function for binary classification to Identify good versus bad particles.

Student intern

• Data analysis Internship: To research unknown adverse of effects of drugs using the FAERS (FDA Adverse Event Reporting System) data base, and find any new and crucial side effects of drugs by writing filter functions on Linux to search for a unique scenario of drug usage.
• Analysing MDMA related deaths: There were nearly 50 deaths with MDMA usage, but after extensive research, we found out that less than 5 were directly related to MDMA.