Sevan Gregory Ficici
sevan AT cs.brandeis.edu
Post-Doctoral Fellow, Harvard University • Ph.D, Brandeis University • Complex Systems Summer School, Santa Fe Institute • MA, Eastman School of Music

Industrial Work • Academic Research

My industrial efforts use methods from a variety of disciplines, such as artificial intelligence, machine learning, combinatorial optimization, constraint systems, data analytics and visualization, and statistics, to build effective (and, where needed, novel) solutions to problems. Below, I highlight and briefly describe some of this work.

Natural Selection

Natural Selection provides consulting and research for privately and federally funded projects. My work here involves a wide variety of problem domains, as well as a wide variety of analytical and machine learning methods.
  • Problem domians include: Bioinformatics, environmental remediation, failure prediction, finance, fraud detection, market research, risk assessment, scheduling
  • Methods include (not exhaustive): Supervised learning (e.g., SVM, ANN, generative models, regression, random forrests), unsupervised learning (e.g., DBSCAN, k-means), visualization (e.g. t-SNE), dimensionality reduction (e.g., PCA), data cleansing, statistical analysis.
  • Languages: Python (with TensorFlow, Keras, SciKit Learn, pandas, numpy, Seaborn); R (with Tidyverse); Java; C

Affinnova

Affinnova was a startup aquired by Nielsen in 2014. As Senior Research Scientist, I was responsible for the design and prototyping of innovative, state-of-the-art, patented interactive evolutionary algorithm and simulation technology that is used to design products and packaging for internationally recognized consumer brands. Affinnova is discussed in the New York Times Best-Seller The Second Machine Age.
  • In addition to designing and prototyping, I contributed to production code
  • Language: Java
Patents

Afeyan, N.B., Malek, K.M, Bufton, N.J, Ficici, S.G., Austin, H.A., Austin, L.J., McClellan, H.E., Method for iterative design of products. United States Patent 7,730,002

Afeyan, N.B., Malek, K.M, Bufton, N.J, Ficici, S.G., Austin, H.A., Austin, L.J., McClellan, H.E., Method and apparatus for evolutionary design. United States Patent 7,610,249

Afeyan, N.B., Malek, K.M, Bufton, N.J, Ficici, S.G., Austin, H.A., Austin, L.J., McClellan, H.E., Method and apparatus for dynamic, real-time market segmentation. United States Patent 7,177,851