Russell Foltz-Smith’s Post

🚀 Excited to share our latest research article, "GeoRF: A Geospatial Random Forest", published in Data Mining and Knowledge Discovery! Proud to highlight Margot Geerts as the lead researcher, with valuable collaboration from Seppe vanden Broucke. In this work, we focus on tailoring state-of-the-art Random Forest models to better handle geographical coordinates as features. While traditional tree-based models typically employ univariate, axis-parallel splits, our approach introduces bivariate splits for simultaneous learning from X- and Y-coordinates. This customization allows our model to generate axis-parallel, oblique, and Gaussian patterns in geographic space, enhancing accuracy to a level comparable to boosting models. Additionally, it improves the interpretability of decision boundaries through more intuitive visualization on geographic maps. Curious to learn more? Check out the full article here: https://lnkd.in/e-djMqpE #Research #DataScience #GeospatialAnalysis #MachineLearning #RandomForest #GIS #HousePricePrediction #RealEstate #Valuation

New discrimination diagrams for basalts based on big data research: In recent days, discrimination diagrams have been widely used for tracing the tectonic settings and origins of basalts from orogenic belts. However, ... #bigdata #cdo #cto

"Generative AI-Berit can you show me the wells on "insert field"?" One of the most crusial point for any engineer is knowing what data you have available and where the data comes from. In this video you see 🔹data availablility overview - out of 194 wellbores, 185 have geology data 🔹Data source overview - which data source did we get the data from, how many wellbores did we get and which wellbores are included. 🔹Pro Well Plan AS have extracted, cleaned and contextualized data from 11 different databases - now also including reports. 🔹You need to trust the data and know where the data comes from. #GenerativeAI

🌿 Exciting News: Unveiling My Pilot Exercise Study on Monthly Consumption Forecasting for Lulu Island using ArcGIS Pro Space-Time Pattern Mining Tool! 🌍📊 Such studies' outcomes are nothing short of fascinating. Uncovering of how consumption behaviors of our consumers evolve over time, enabling stakeholders to make informed decisions for resource allocation, sustainable development, and economic growth. I'm eager to hear your thoughts and insights on my exercise. Let's share how geospatial data and #GIS #ESRI #ARCGISPRO can drive sustainable growth. #ArcGISPro #SpaceTimePatternMining #DataAnalysis #Geospatial #LuluIsland #ConsumptionForecasting #Sustainability #EconomicDevelopment

Rock vs Mine Prediction Project using Logistic Regression #dataanalysis #machinelearning #machinelearningalgorithms Here is the Repo https://lnkd.in/eUNhFXcU

Unlocking the Power of Genetic Algorithms in Data Mining ▸ https://lttr.ai/AIJwh #VastLandscape #DataAnalysis #GeneticAlgorithms #PowerfulTools #HiddenInsights #General #CommonIssueFaced #ClassifyingInstancesCorrectly #ReduceFalsePositives #DetectFraudulentActivities

Variogram Modeling. Variogram calculation Process Calculate the Experimental Variogram. Fit a Variogram Model to the Experimental Variogram. Variogram model types could be Spherical, Gaussian, or Exponential. Interpretative Process Take into account geological knowledge. Vertical Variogram Model Usually plenty of data and easily estimated. Horizontal Variogram Model Cannot be calculated due to limited amount of data. Can be derived from correlated data source or taken from analogous field / outcrop / geological knowledge.

In the vast landscape of data analysis, genetic algorithms and data mining stand tall as powerful tools that unlock the hidden insights within complex... Read more 👉 https://lttr.ai/AHqq2 #VastLandscape #TreasureTrove #InsightsWaiting #General #CloselyIntertwinedFields #ArchaeologistSMeticulousPreparation #ContinuousTargetVariable #BookSOverarchingTheme #IntrepidExplorersVenturing #MaintainRelativeDistances

Best sulfide segmentation technique I have found. Subtract two images, one taken with cross polarized light and the other with plane polarized light. The method can also be augmented with a machine learning model to further segment the sulfide species, for example Chalcopyrite vs. Pyrite. Could be used when logging, geologist could take the images and and app could return percentages within field of view. Useful?

Geological machine learning model of gold across Africa (surface model): https://lnkd.in/eWZqfmkw