Associate Data Scientist - Environmental Modeling

Associate Data Scientist – Environmental Modeling

Associate Data Scientist – Environmental Modeling for Chesterfield, MO to design & build statistical, machine learning and deep learning models to quantify subfield-scale yield testing environments of crops; automate analytics workflows; develop next generation methodologies for integrative usage of genomic, phenomic & environmental data; determine environmental correlations among testing locations & global regions; design statistical modeling frameworks & prediction models to drive product placement recommendations and yield predictions; collaborate to provide data-driven statistical solutions to business problems. Requires Master’s in Statistics, Mathematics, or closely related quantitative field & 1 yr experience using object-oriented programming techniques to write Python packages to analyze high dimensional environmental data with Gap Statistics; developing & selecting unsupervised learning algorithms to analyze high-dimensional environmental data, including K-means, agglomerative hierarchical clustering, and/or Gaussian mixture models; using statistical & machine learning packages, including Tensorflow, Pandas, Multiprocessing, Joblib, Numpy, SciPy, Scikit-Learn, Keras, PyTorch, PySpark, and/or Dask, to develop discovery and production ready models for analysis of phenotypic and geospatial data; adhering to and/or enforcing coding best practices; using code management tools, including GitHub, to ensure the reproducibility of data science; aggregating & summarizing complex datasets using GCP BigQuery, Presto, Superset, and AWS RedShift; building heat, drought, and cold stress models over global regions using high dimensional environmental data; automating workflows using AWS Sagemaker, Google Cloud Platform, Airflow, & Docker; performing data operations, including spatial joins, zonal statistics, & re-projecting; quantifying similarity scores between different environments & using distance metrics to compare multivariate time series environmental data related to major row crops; visualizing geospatial data, including vector & raster files, using QGIS, Google BigQuery, and/or Python libraries; performing data quality checks using deep learning-based anomaly detection on time-series data; and designing, training & optimizing neural networks for generating embeddings using AutoEncoder for multivariate time series-based data.  Position may telecommute from home office location within reasonable commuting distance of Chesterfield, MO up to 4 days per wk.  Mail resume to Cascinda Fischbeck, Bayer Research and Development Services, LLC, 800 N. Lindbergh Blvd. E2NE, St. Louis, MO 63167 or email resume to careers_us@bayer.com.  Include reference code below with resume.