Data Science Approaches

Data science empowers managers to bring their information to bear on their projects.  NEI uses several branches of theoretical and applied data science to assist in decision making.  These include:

Management of large data sets

NEI specializes in accessing, cleaning, and integrating large public data sources into the biological data sets collected for or by our clients.

Information theoretic techniques

NEI uses information theoretic methods to build quantitative predictive models.  Time and again, decision-makers find that the evidence ratios produced by this approach are straightforward, transparent, and compelling in the predictive evaluation of risk.  NEI works closely with decision makers to build detailed models based on the most relevant features of a given system.

Fisherian frequentist statistics

Alpha and p-value-based analysis approaches are legacy items of 1930s computing power.  NEI is available for consulting regarding these methods, including t-tests, ANOVA, linear and non-linear regression, and Chi-squared statistics.  Where possible, we are also providing clients with more current options.

Deterministic models

This class of analyses fits data to specific mathematical models for which real-world deviation from the model is expected to be minimal.

Survival analysis

Also termed time-to-event and time-to-failure analysis, NEI develops a model best suited to each data set, often through the Cox proportional hazards regression technique.

The R language environment

In the past decade, the R project has emerged as the most powerful environment for analysis, and is the data workhorse of NEI.  Clients are provided with annotated copies of analyses, and professional-grade outputs in their preferred format.