Gerald Harp, Ph.D. Physics Email & Phone Number

• Platforms and Languages: OpenMP, Java, Python, Pytorch, C, C++, some Ruby, some SQL, MathLab, Mathematica, Threading Building Blocks, FORTRAN, Linux, Windows. Domains and Expertise: Machine learning and Automation.
• Expert consultant, prototyping lunar interferometry proof of principle for Principia College Institute.
• Panel reviewer, National Science Foundation, to provide guidance on the next-generation processing systems to support.
• External technical reviewer on a panel member at Jet Propulsion Laboratory.
• External technical reviewer on a panel member at Very Large Array Radio telescope.
• Expert consultant on application to troubleshoot magnetism issue for ABB (ASEA Brown Boveri).

Basking Ridge, New Jersey, US

Machine learning using Numpy, Pandas, Sklearn making Predictive Models of suicide risk. Isolated 80% of adults classified as high risk into only 20% of full population.Deep learning using Pytorch and creating a Limited Boltzmann Machine (i.e. Neural Network) -- subject matter is classified.Natural Language Processing study of female sentiments of.

• Discriminative classification with nonlinear regression, logistic regression, tree models (XGBoost), and convolutional neural networks.
• Applications to adapt data with unbalanced minority/majority class datasets, working with privacy-sensitive health care and behavioral data using techniques such as neural networks and supervised or unsupervised.
• Fitted parameter confidence levels were determined analytically or via multi-fold cross-validation. Optimization as needed.
• Realistic synthetic data generation with generative joint-probability distributions, local linear embedding, neural networks using autoencoders (VAEs), and data generative adversarial networks (GANs).
• Time series analysis for pattern discovery and extrapolation to future times. Analysis of multi-sensor data with disparate data acquisition rates.
• Noise suppression with Fourier methods, principal/independent component decomposition, artifact elimination via projections of known signal types and pattern matching.

Mountain View, California, US

• Director of SETI Research (12-18), Senior Astrophysicist (10-12), Astrophysicist (09-10), Senior Software Scientist (00-09). Research with radio-frequency signals. Design and implement experimental tests of instruments.
• Partnering with IBM at SF venue, led Hackathon / Code Challenge to develop neural networks that discover intelligent signals buried in astronomical data, i.e. the search for extraterrestrial intelligence.
• Created Deep Convolutional Neural Networks (CNNs) for High-Fidelity Images from Radio Interferometer Data (e.g. detected Fast Radio Bursts).
• Simulated radio signals using Wide Residual Neural Networks with Machine Vision and Deep Learning to classify signals.
• Led two successful Air Force research projects for satellite astrometry and as a receiver in a bistatic radar space debris tracking. Accomplished on time and met goals, working with teams of direct and dotted line.
• Led dozens of studies using a variety of techniques (e.g., polarization squint, dish surface characterization, beam analysis, testing ground shields).

Project leader with direct reporting staff on multiple contracts, including those for DoD and NEC Japan. Led team of young software engineers in using large datasets for the development of a prototype virtual reality simulator for visualization of 2D and 3D object motion modeling and use in military flight simulators and commercial game-style simulators.

Athens, Ohio, US

• As professor of physics at Ohio University, awarded prestigious NSF Career Grant 1996. From NSF home website: “The Faculty Early Career Development (CAREER) Program … offers the National Science Foundation's most.
• Built state-of-the-art lab comprising a custom vapor deposition system, novel electron diffraction and photoemission instrumentation, x-ray diffraction, and magnetometers for ex-situ analysis.
• Designed and built several novel research instruments including, an ultra-high vacuum cryogenically pumped sputter deposition system to create 2D atomic structures and study magnetic properties of artificial atomic.
• Updated hardware and added automation to magneto-optical Kerr magnetometer. Updated and automated a 2-axis x-ray diffractometer for use with single-crystal films. Built multiple synchrotron-based UHV systems to perform.
• Served as senior advisor for 4 successful Ph. Ds and 1 master’s degree student who are all presently employed in fields relating to their research.
• Achieved tenure in record time (five years).

Armonk, New York, NY, US

• Worked on magnetism and magnetic properties through control of crystal structure in ultrathin metallic superlattices.
• Developed new methods to prepare magnetic/non-magnetic superlattices in metastable crystal structures and study their magnetic, structural, and electronic properties.
• Pioneered the use of polarized x-ray absorption spectroscopy to study physical and electronic structure effects on magnetism in such materials.
• Rebuilt and automated dual-axis angle-resolved photoelectron spectrometer. Wrote numerous peer-reviewed publications with colleagues research scientists.
• Developed a New Chemically Ordered Alloy characterized through Magneto-Optical Kerr Spectroscopy.
• Studied the Growth Temperature Dependence of Magnetoresistance Superlattice Wedges.

M.S. Physics, B.S. Physics, B.S. Mathematics, Minor Chemistry

Doctor Of Philosophy - Phd, Physics