Cynthia Freeman - Academia.edu (original) (raw)

Uploads

Papers by Cynthia Freeman

Journal of Artificial Intelligence Research, Nov 18, 2021

The existence of an anomaly detection method that is optimal for all domains is a myth. Thus, the... more The existence of an anomaly detection method that is optimal for all domains is a myth. Thus, there exists a plethora of anomaly detection methods which increases every year for a wide variety of domains. But a strength can also be a weakness; given this massive library of methods, how can one select the best method for their application? Current literature is focused on creating new anomaly detection methods or large frameworks for experimenting with multiple methods at the same time. However, and especially as the literature continues to expand, an extensive evaluation of every anomaly detection method is simply not feasible. To reduce this evaluation burden, we present guidelines to intelligently choose the optimal anomaly detection methods based on the characteristics the time series displays such as seasonality, trend, level change concept drift, and missing time steps. We provide a comprehensive experimental validation and survey of twelve anomaly detection methods over different time series characteristics to form guidelines based on several metrics: the AUC (Area Under the Curve), windowed F-score, and Numenta Anomaly Benchmark (NAB) scoring model. Applying our methodologies can save time and effort by surfacing the most promising anomaly detection methods instead of experimenting extensively with a rapidly expanding library of anomaly detection methods, especially in an online setting. 1. Similar analysis (Emmott et al., 2015) has been performed before to compute the influence of meta-data on anomaly detection but for feature-vector-based datasets instead of time series. 2. See Appendix for a table of all acronyms and their definitions. 3. See https://github.com/dn3kmc/jair anomaly detection for all source code implementations, Jupyter notebooks demonstrating how to determine characteristics, and datasets.

Massachusetts, New Jersey, Connecticut, and Maine have adopted regulatory zones around seasonal (... more Massachusetts, New Jersey, Connecticut, and Maine have adopted regulatory zones around seasonal (vernal) pools to conserve terrestrial habitat for pool-breeding amphibians. Most amphibians require access to distinct seasonal habitats in both terrestrial and aquatic ecosystems because of their complex life histories. These habitat requirements make them particularly vulnerable to land uses that destroy habitat or limit connectivity (or permeability) among habitats. Regulatory efforts focusing on breeding pools without consideration of terrestrial habitat needs will not ensure the persistence of poolbreeding amphibians. We used GIS to combine a discrete-choice, parcel-scale economic model of land conversion with a landscape permeability model based on known habitat requirements of wood frogs (Lithobates sylvaticus) in Maine (USA) to examine permeability among habitat elements for alternative future scenarios. The economic model predicts future landscapes under different subdivision open space and vernal pool regulatory requirements. Our model showed that even "no build" permit zones extending 76 m (250 ft) outward from the pool edge were insufficient to assure permeability among required habitat elements. Furthermore, effectiveness of permit zones may be inconsistent due to interactions with other growth management policies, highlighting the need for local and state planning for the long-term persistence of pool-breeding amphibians in developing landscapes.

Journal of Artificial Intelligence Research, Nov 18, 2021

The existence of an anomaly detection method that is optimal for all domains is a myth. Thus, the... more The existence of an anomaly detection method that is optimal for all domains is a myth. Thus, there exists a plethora of anomaly detection methods which increases every year for a wide variety of domains. But a strength can also be a weakness; given this massive library of methods, how can one select the best method for their application? Current literature is focused on creating new anomaly detection methods or large frameworks for experimenting with multiple methods at the same time. However, and especially as the literature continues to expand, an extensive evaluation of every anomaly detection method is simply not feasible. To reduce this evaluation burden, we present guidelines to intelligently choose the optimal anomaly detection methods based on the characteristics the time series displays such as seasonality, trend, level change concept drift, and missing time steps. We provide a comprehensive experimental validation and survey of twelve anomaly detection methods over different time series characteristics to form guidelines based on several metrics: the AUC (Area Under the Curve), windowed F-score, and Numenta Anomaly Benchmark (NAB) scoring model. Applying our methodologies can save time and effort by surfacing the most promising anomaly detection methods instead of experimenting extensively with a rapidly expanding library of anomaly detection methods, especially in an online setting. 1. Similar analysis (Emmott et al., 2015) has been performed before to compute the influence of meta-data on anomaly detection but for feature-vector-based datasets instead of time series. 2. See Appendix for a table of all acronyms and their definitions. 3. See https://github.com/dn3kmc/jair anomaly detection for all source code implementations, Jupyter notebooks demonstrating how to determine characteristics, and datasets.

Massachusetts, New Jersey, Connecticut, and Maine have adopted regulatory zones around seasonal (... more Massachusetts, New Jersey, Connecticut, and Maine have adopted regulatory zones around seasonal (vernal) pools to conserve terrestrial habitat for pool-breeding amphibians. Most amphibians require access to distinct seasonal habitats in both terrestrial and aquatic ecosystems because of their complex life histories. These habitat requirements make them particularly vulnerable to land uses that destroy habitat or limit connectivity (or permeability) among habitats. Regulatory efforts focusing on breeding pools without consideration of terrestrial habitat needs will not ensure the persistence of poolbreeding amphibians. We used GIS to combine a discrete-choice, parcel-scale economic model of land conversion with a landscape permeability model based on known habitat requirements of wood frogs (Lithobates sylvaticus) in Maine (USA) to examine permeability among habitat elements for alternative future scenarios. The economic model predicts future landscapes under different subdivision open space and vernal pool regulatory requirements. Our model showed that even "no build" permit zones extending 76 m (250 ft) outward from the pool edge were insufficient to assure permeability among required habitat elements. Furthermore, effectiveness of permit zones may be inconsistent due to interactions with other growth management policies, highlighting the need for local and state planning for the long-term persistence of pool-breeding amphibians in developing landscapes.