The k-nearest neighbor (k-NN) technique is a non-parametric technique that can be used to make predictions of discrete (class-type) as well as continuous variables. The k-NN technique and many of its derivatives belong to the group of .lazy learning algorithms.. It is lazy, as it passively stores the development data set until the time of application; all calculations are performed only when estimations need to be generated.
Irrigator Pro is an expert system designed to provide irrigation scheduling recommendations based on scientific data resulting in conservation minded irrigation management. The success of Irrigator Pro for Peanuts created interest in other groups. A collaborative effort between the NPRL, Cotton Commission, University of Georgia, and the Peanut Foundation was established to create comparable models for cotton and corn.
The WeedCast Software Suite is a decision aid that currently includes the WeedCast and WheatScout models. These models were created by the USDA ARS and the University of Wisconsin. The software is written in Java and is free to download and use. The source code is released under the GPL.
SPUR2 DOS ver. 2.2 is a general grassland ecosystem simulation model designed to determine beef cattle performance and production by simultaneously simulating production of up to 15 plant species on 36 heterogeneous grassland sites. SPUR2 simulates grassland hydrology, nitrogen cycling, and soil organic matter on grazed ecosystems as well as rangeland production under different climatic regimes, environmental conditions, and management alternatives.
Root Zone Water Quality Model 2 (RZWQM2) is a whole-system model for studying crop production and environmental quality under current and changing climate conditions. It emphasizes the effects of agricultural management practices on physical, chemical and biological processes. RZWQM2 is a one-dimensional model with a pseudo 2-dimensional drainage flow. Crop simulation options include the generic plant growth model, DSSAT-CSM 4.0 and HERMES SUCROS models. It also can simulate surface energy balance with components from the SHAW model and water erosion from the GLEAMS model. An automated parameter estimation algorithm (PEST) was added to RZWQM2 for objective model calibration and uncertainty analysis.
PhenologyMMS is a simulation model that outlines and quantifies the developmental sequence of different crops under varying levels of water deficits, provides developmental information relevant to each crop, and is intended to be used either independently or inserted into existing crop growth models.
An interactive computer program was developed to simulate the interactions among the above factors. It shows how changing one factor impacts the outcome of the other factors for a single growing season. The user selects a climate, a crop, and soil characteristics from menu lists, and then sets the water table depth and quality, irrigation (river or well) water quality and then develops an irrigation schedule. On execution, the relative yield reductions due to over irrigation, under irrigation, and salinity, water table rise or fall and surface runoff are shown numerically for the growing season. Soil water content, soil salinity, water table depth changes and rain and irrigation events during the season are also shown graphically.
Predicts the salinity, sodicity, and toxic-solute concentration of the soil-water within a simulated crop root zone resulting from the use of a particular irrigation water of given composition and at a specified leaching fraction. It can be used to evaluate the effect of a given salinity level (or solute concentration) on crop yield and of a given sodicity level on soil permeability.