Annual International Conference On GIS, GPS AND Remote Sensing.

Annual Annual

International Conference

On GIS , GPS AND Remote Sensing

R.Naveen Kumar Goud

B.E 4/4 Civil Engineering

Vasavi College Of Engineering

Hyderabad, Andhra Pradesh ,India

GA Optimization Technique OnGA Optimization Technique On

Spatio-Temporal Interpolation Spatio-Temporal Interpolation

For Dynamic GIS For Dynamic GIS

abstractabstract • Generating cross-sectional data of

arbitrary time slice is a basic function to support temporal operations such as time-series analysis and integration of dynamic models in GIS environment.

• Here we propose a interpolation scheme for class variable data under the framework of optimization of likelihood.

Objective of the Study Objective of the Study

• Dynamic analysis of spatial data are needed in various fields .

• Difficult to generate spatio-temporal filed of quality data for analysis .

• Quality data, models describing structure is integrated with observational data .

• Integration methods for data and models have been mainly developed for continuous variables in meteorology and oceanography .

• For class variables such as land use types, there are primitive interpolation methods,nearest neighbor interpolation .

Genetic AlgorithmGenetic Algorithm

• GA are used as approach to optimization which requires efficient and effective search.

• They combine survival of the fittest among structures .

five basic aspects five basic aspects

• the representation of problem,

• the initialization of population,

• the definition of evaluation function,

• the definition of genetic operators,

• the determination of parameters .

Optimization Scheme for Optimization Scheme for Variable InterpolationVariable Interpolation

• S-T data can be divided into two types: continuous VD & class VD .

• The estimation of time of changes according to "class boundary distance".

• we go to integrate observational class data with structural models to make robust and reliable spatio-temporal interpolation of class variables.

• Searching for spatio-temporal field of class

data is typical combinatorial optimization problem, we introduce the genetic algorithm as a optimization scheme.

• The likelihood is computed based on both the fitness to observational data and that to behavioral models.

Application of Genetic AlgorithmApplication of Genetic Algorithm

• 3D Representation of an Individual.

• Initialization of Population .

• Definition of Individual's Fitness .

Representation of an IndividualRepresentation of an Individual

The S-T relations affect the The S-T relations affect the transitional probability in three waystransitional probability in three ways

• Spatial Continuity: Assumption that the same class data tends to continue in spatial dimension.

• Temporal Continuity: This is an extension of spatial continuity to temporal domain.

• The third aspect is Expansion-Contraction relations:

Definition of Individual's FitnessDefinition of Individual's Fitness

Calculation of Individual's FitnessCalculation of Individual's Fitness

• Individual's fitness has two parts:

behavioral fitness & observational fitness.

• By multiplying behavioral fitness and observational fitness, overall fitness can be computed.

• To integrate behavioral models and observational data, the overall fitness has to be optimized.

Definition of OperatorsDefinition of Operators

Reproduction :

• This is a process in which individual strings are copied according to their objective function values or the fitness values

• That strings with a higher value have a higher probability of contributing one or more offspring in the next generation

ReproductionReproduction

CrossoverCrossover

MutationMutation

• Mutation operator plays a secondary role in the simple GA.

Hill-Climbing methodHill-Climbing method

• That exploits the best among known possibilities for finding an improved solution.

• Although Hill-Climbing strategies is easy to trap in one of local maxima more far away from the optimal solution.

Conclusion and Future ProspectsConclusion and Future Prospects

• GA/HC can be very rigorous because it can generate the most likely spatio-temporal distribution of class variables under observational data and a behavioral model.

• Hill-Climbing method can be effective method to greatly improve the efficiency of GA.

ReferencesReferences • Bramlette, M.F. (1991): Initialization, Mutation and Selection

Methods in Genetic Algorithms for Function Optimization, • Davis, L. (1987) : Genetic Algorithms and Simulated Annealing, • Eshelman, L.J. and J.D.Schaffer (1991): Preventing Premature

Convergence in Genetic Algorithms by Preventing Incest. • Goldberg, D.E. (1989) : GENETIC ALOGRITHMS in Search,

Optimization and Machine Learning. • Gold, C.M.(1989): Surface interpolation, spatial adjacency and GIS,

Three Dimensional Applications in Geographic Information System, • Huang, S.B. and R.Shibasaki(1995): Development of Genetic

Algorithm /Hill-climbing Method for Spatio-temporal Interpolation,. • Shibasaki,R., T.lto and Y.Honda (1993):• http://www.oursland.net/projects/PopulationExperiment/

Thank You