This tool is actually embeded in the TBMod component. It allows a user to adjust “user-friendly” parameters for the Bayesian model and see real-time feedback of how parameter changes affect the final probability distribution generated by the model. User can choose to use recommende parameters (from past statistics) or use own parameters. User can also specify whether to use human behavior data or not (get posterior prediction or not).

Input Parameters:

• Terrain Features (Only existing ones for given search region) ←- Provided by TBMod in hex coordinates matrices)
  • Topography (lake, hill, plane, etc.)
  • Vegetation density (Sparse, medium, dense)
  • Elevation (Uphill, flat, downhill)
  • Trail (On trail, off trail)
• Wether intended destination known? If so, hex coordinates.
• Wether trails exist
• Max duration (e.g. 2 hours since missing person last seen)
• Prior predictive distribution (Array of matrices)
• (Optionally) Posterior predictive distribution (Array of matrices)

Output Parameters:

• Prior beliefs
  • Transitional probability topography: Alpha and Beta values for Beta distributions
  • Transitional probability vegetation density: Alpha and Beta values for Beta distributions
  • Transitional probability elevation: Alpha and Beta values for Beta distributions
  • Transitional probability staying on trail: Alpha and Beta value for Beta distribution
  • Transitional probability intended destination: mean and standard deviation for Gaussian distribution

• Interface to specify missing person profile
  • Age: kids, adults, elderly
  • Gender: Male, female
  • Type: Scouts, hikers, hunters, snowshoers, etc.
  • Season: Winter, Spring, Summer, Fall
  • Weather: Sunny/Cloudy, Raining, Snowing, etc.
• Checkbox for using recommended parameters (from past statistics)
• Generate look up table/database of parameters for different profiles
• Sliders for adjusting mean and standard deviation.
• Create tool to compute Alpha and Beta values for Beta distribution from means and standard deviations.
• Draw current Beta distribution curve based on given Alpha and Beta values.
• Checkbox for whether to use GPS track log data.
• Requesting TBMod for Prior Predictive and Posterior Predictive distributions.
• Painting distribution map with color map (hot areas vs cold areas)
• Store already computed distribution map data (to avoid expensive computation)
• Play animation of distribution change over time and allow user to pick a frame (only repaint this frame when adjusting parameters)
• Radio button to switch compare Prior Predictive vs. Posterior Predictive distributions
• Button to accept generated distribution.
• Send accepted generated distribution to repository.