Machine Learning Tools in ArcGIS
Machine learning isn’t science fiction, but instead consists of a series of data-driven algorithms and techniques that automate the prediction, classification and clustering of data. Spatial machine learning incorporates geography into calculations, as well as the shape, density, contiguity, spatial distribution, and proximity as it relates to the data.
Prediction
When machine learning algorithms make predictions, they’re using the known to estimate the unknown. Walk through an entire suitability analysis of Cougar habitats, built to address specific criteria that cater to Cougar habitat including steep slopes, forested areas, proximity to streams, distance from major roads and highways, and the presence of prey like deer or elk.
A variety of tools available for prediction modeling in ArcGIS include: Empirical Bayesian Kriging, Areal Interpolation, EBK Regression Prediction, Ordinary Least Squares Regression and Exploratory Regression, Geographically Weighted Regression, and Forest-based Classification and Regression.
Classification
This is the process by which a given object is classified into a category based on a trained dataset. Decision trees build classification models that utilize an if-then rule, which are learned sequentially using trained data over time.
Density-based Clustering & Forest-based Classification and Regression – Video from esri
Tools in ArcGIS include: Maximum Likelihood Classification, Random Trees, Support Vector Machine, and Forest-based Classification and Regression.
Clustering
Clustering is a grouping of observations based on similarities of values or locations in the dataset. This unsupervised learning method is essentially a collection of objects on the basis of similarities and dissimilarities between the data.
Tools in ArcGIS include: Spatially Constrained Multivariate Clustering, Multivariate Clustering, Density-based Clustering, Image Segmentation, Hot Spot Analysis, Cluster and Outlier Analysis, and Space Time Pattern Mining.
ArcGIS Integrates with External Frameworks
As an open, interoperable platform, ArcGIS enables the integration of complementary methods including ArcGIS API for Python, ArcPy, and the R-ArcGIS Bridge. With this integration, ArcGIS users can approach and solve complex spatial problems by combining these powerful built-in tools with any machine learning package. This includes scikit-learn and TensorFlow in Python, to caret in R or ITM Watson and Microsoft AI, which all benefit from the visualization capabilities, spatial validation, and geoenrichment in ArcGIS.
Learn more about the R-ArcGIS Bridge, or dive right in with a hands-on lesson to Analyze Crime Using Statistics and the R-ArcGIS Bridge.
Video from esri
What’s Next?
Esri has a number of key initiatives to advance both machine learning and deep learning approaches, with integration throughout the platform and a major focus on big data and distributed computing with Spark ML.
Ready to learn more about machine learning techniques within the ArcGIS platform? Explore a few esri resources to dive deeper and continue to experience the power of algorithms. Visit: http://desktop.arcgis.com/en/analytics/casestudies/ or https://spatialstats.github.io/
Get Support
Are you seeking professional geospatial services that incorporate machine learning? Then you’ve come to the right place. Reach out to our industry experts to start a conversation today, and let’s begin discussing your unique needs.
