AU2021103651A4

AU2021103651A4 - A method and system for performing weather forecast

Info

Publication number: AU2021103651A4
Application number: AU2021103651A
Authority: AU
Inventors: G. Kavitha; S. Krithika; Desai Manish
Original assignee: Hindustan Institute Of Technology & Science Deemed To Be Univ
Current assignee: Hindustan Institute Of Technology & Science Deemed To Be University
Priority date: 2021-06-26
Filing date: 2021-06-26
Publication date: 2022-03-31
Anticipated expiration: 2029-06-26

Abstract

The present invention generally relates toa method and system for performing weather forecast. The method comprises partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector; initializing weather parameters in first phase using interval values Markov integrated Rhotrix predictive model; and optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling. 17 100 Band Pass Filter2 Markov integrated Rhotrixpredictive model104 OptimizationUnit CentralProcessing 106 Unit 108 Figure 200 partitioningcontinuousnumericaldatasetvariablestodiscretize dintervalsofequal lengths forming an interval valued partitioned vector A -204 initializing weather parameters infirst phase using intervalvalues Markov integrated 2 R trix predictivemodel optimizingestimationofweatherparametersinsecondphaseusinggeneticapproach 206 for reducing error in predictive modeling Figure

Description

Band Pass Filter2 Markov integrated Rhotrixpredictive model104

OptimizationUnit CentralProcessing 106 Unit 108

Figure

200

partitioningcontinuousnumericaldatasetvariablestodiscretize dintervalsofequal lengths forming an interval valued partitioned vector A -204 initializing weather parameters infirst phase using intervalvalues Markovintegrated 2 R trix predictivemodel

optimizingestimationofweatherparametersinsecondphaseusinggeneticapproach 206 for reducing error in predictive modeling

Figure

A METHOD AND SYSTEM FOR PERFORMING WEATHER FORECAST FIELD OF THE INVENTION

The present invention relates to a method and system for performing weather forecast.

BACKGROUND OF THE INVENTION

Predictive modeling is a learning approach based on data to create a robust accurate model to make predictions. Uncertainty and Randomness lead to high fluctuations making it difficult for decision makers to make decisions. In today's world, both structured and unstructured publicly available data from open data sources are too huge and complex to analyze and extract information for building a model, generating the result, optimizing the parameters and validating the results.

A Markov chain is an essential part of stochastic process satisfying the Markov ageless property which means that knowing the current state of the process, future prediction can be done in a best possible way with no background of its past states. The model proposed aims at partitioning the states, determining the transition probabilities and long run probability vector.

A special form of matrix called coupled matrix is integrated with Markov to solve the problem involving n x n state transitions and (n-1) x (n-1) state transitions simultaneously. Using coupled rhotrix, an integrated Markov Rhotrix model is developed for the state matrix.

Genetic approach is an optimization technique for optimizing the unknown parameters. The Initial set of parameter population selected is used to compute the fitness of the function. On repeated selection process using crossover and mutation operators of genetic approach the fitness is computed till optimal population convergence is reached. The proposed model involves Genetic Approach for Markov integrated Rhotrix optimization.

In the view of the forgoing discussion, it is clearly portrayed that there is a need to have a method and system for performing weather forecast. SUMMARY OF THE INVENTION

The present disclosure seeks to provide a method and system for developing an integrated model in a novel way for weather forecasting application.

In an embodiment, a system for performing weather forecastis disclosed. The system includes a pre-processing unit for partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector. The system further includes a Markov integrated Rhotrix predictive model for initializing weather parameters interval values in first phase. The system further includes an optimization unit for optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

In an embodiment, the central processing unit is configured with the interval valued Markov integrated Rhotrix predictive model for initializing weather parameters.

In another embodiment, a method for performing weather forecast is disclosed. The method includes partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector. The method further includes initializing weather parameters in first phase using interval values Markov integrated Rhotrix predictive model. The method further includes optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

In an embodiment, the dataset vector determine minimum, maximum and range choosing number of equal subintervals to be partitioned.

In an embodiment, the interval valued partitioned vector formed is having finite set of points according to the number of subintervals chosen.

In an embodiment, modelling of Rhotrix predictive model comprises calculating two matrices by Markov Chain transition probability for forming the Rhotrix and then integrating it with Rhotrix for predictive modelling; calculating Markov chain integrated Rhotrix transition probability and steady state probability in the long run for combined impact of features on selected for prediction; and optimizing unknown parameters for combined impact of features and thereby entering genetic optimization procedure in the next phase for predictive modeling.

In an embodiment, steps for optimizing GA comprises loading input dataset thereby partitioning into equal sub intervals; calculating Markov chain transition probability of (nxn) dimension and steady state probability in the long run of (lxn) dimension for a first feature selected for prediction; calculating Markov chain transition probability of {(n 1)x(n-1)} dimension and steady state probability in the long run of

[lx(n-1)] dimension for a second feature selected for prediction; calculating Markov chain integrated Rhotrix transition probability of {[n+(n-1)]x[n+(n-1)] dimension and steady state probability in the long run of {1x[n+(n-1)]} dimension for combined impact of the second feature on first feature selected for prediction; applying GA approach for parameter optimization of first case, second case, and third case as and thereby performing training and testing on the entire dataset using the Genetic approach operators and thereafter error measures are computed; comparing result obtained with the actual output using GA fitness function; and terminating optimization when stopping criteria, the average change in the fitness value less than options is reached and thereby displaying the predictive modeling output.

In an embodiment, there are (nxn) unknown parameters are optimized for the first feature, considered as first case, whereas there are {(n-1)x(n-1)} unknown parameters are optimized for the second feature, considered as second case.

In an embodiment, there are {[n+(n-1)]x[n+(n-1)] unknown parameters to be optimized for combined impact of second feature on first feature, which is considered as third case.

In an embodiment, the genetic operators used are selection, crossover and mutation operators.

An object of the present disclosure is to develop of a new predictive modelling based on interval valued Markov integrated Rhotrix optimization using genetic approach.

Another object of the present disclosure is to facilitate short and decadal long weather prediction.

Yet another object of the present invention is to develop expeditious and cost-effective method for performing weather forecast.

To further clarify advantages and features of the present disclosure, a more particular description of the invention will be rendered by reference to specific embodiments thereof, which is illustrated in the appended drawings. It is appreciated that these drawings depict only typical embodiments of the invention and are therefore not to be considered limiting of its scope. The invention will be described and explained with additional specificity and detail with the accompanying drawings. BRIEF DESCRIPTION OF FIGURES

These and other features, aspects, and advantages of the present disclosure will become better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:

Figure 1 illustrates a block diagram of a system for performing weather forecastin accordance with an embodiment of the present disclosure; and Figure 2 illustrates a flow chart of a method for performing weather forecastin accordance with an embodiment of the present disclosure.

Further, skilled artisans will appreciate that elements in the drawings are illustrated for simplicity and may not have necessarily been drawn to scale. For example, the flow charts illustrate the method in terms of the most prominent steps involved to help to improve understanding of aspects of the present disclosure. Furthermore, in terms of the construction of the device, one or more components of the device may have been represented in the drawings by conventional symbols, and the drawings may show only those specific details that are pertinent to understanding the embodiments of the present disclosure so as not to obscure the drawings with details that will be readily apparent to those of ordinary skill in the art having benefit of the description herein.

DETAILED DESCRIPTION

For the purpose of promoting an understanding of the principles of the invention, reference will now be made to the embodiment illustrated in the drawings and specific language will be used to describe the same. It will nevertheless be understood that no limitation of the scope of the invention is thereby intended, such alterations and further modifications in the illustrated system, and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates.

It will be understood by those skilled in the art that the foregoing general description and the following detailed description are exemplary and explanatory of the invention and are not intended to be restrictive thereof.

Reference throughout this specification to "an aspect", "another aspect" or similar language means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present disclosure. Thus, appearances of the phrase "in an embodiment", "in another embodiment" and similar language throughout this specification may, but do not necessarily, all refer to the same embodiment.

The terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process or method that comprises a list of steps does not include only those steps but may include other steps not expressly listed or inherent to such process or method. Similarly, one or more devices or sub-systems or elements or structures or components proceeded by "comprises...a" does not, without more constraints, preclude the existence of other devices or other sub-systems or other elements or other structures or other components or additional devices or additional sub-systems or additional elements or additional structures or additional components.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The system, methods, and examples provided herein are illustrative only and not intended to be limiting.

Embodiments of the present disclosure will be described below in detail with reference to the accompanying drawings.

Referring to Figure 1, a block diagram of a system for performing weather forecastis illustrated in accordance with an embodiment of the present disclosure. The system 100 includes a pre-processing unit 102 is configured for partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector.

In an embodiment, a Markov integrated Rhotrix predictive model 104 is conncted with the pre-processing unit 102for initializing weather parameters interval values in first phase. In an embodiment, an optimization unit 106is connected with the predictive model 104for optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

Genetic approach based optimized models are built in less time and highly significant for its global search. Combining the Genetic approach with Interval valued Markov integrated Rhotrix, its implementation and analysis had given a key asset for predictive modeling in weather forecast.

Figure 2 illustrates a flow chart of a method for performing weather forecast in accordance with an embodiment of the present disclosure. At step 202, the method 200 partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector.

At step 204, the method 200 initializing weather parameters in first phase using interval values Markov integrated Rhotrix predictive model.

At step 206, the method 200 optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

In an embodiment, modelling of Rhotrix predictive model including calculating two matrices by Markov Chain transition probability for forming the Rhotrix and then integrating it with Rhotrix for predictive modelling. Then, calculating Markov chain integrated Rhotrix transition probability and steady state probability in the long run for combined impact of features on selected for prediction. Then, optimizing unknown parameters for combined impact of features and thereby entering genetic optimization procedure in the next phase for predictive modeling.

In an embodiment, steps for optimizing GA includes loading input dataset thereby partitioning into equal sub intervals. Then, calculating Markov chain transition probability of (nxn) dimension and steady state probability in the long run of (lxn) dimension for a first feature selected for prediction. Then, calculating Markov chain transition probability of {(n 1)x(n-1)} dimension and steady state probability in the long run of

[lx(n-1)] dimension for a second feature selected for prediction. Then, calculating Markov chain integrated Rhotrix transition probability of {[n+(n-1)]x[n+(n-1)] dimension and steady state probability in the long run of {1x[n+(n-1)]} dimension for combined impact of the second feature on first feature selected for prediction. Then, applying GA approach for parameter optimization of first case, second case, and third case as and thereby performing training and testing on the entire dataset using the Genetic approach operators and thereafter error measures are computed. Then, comparing result obtained with the actual output using GA fitness function. Thereafter, terminating optimization when stopping criteria, the average change in the fitness value less than options is reached and thereby displaying the predictive modeling output.

In an embodiment, consider, a Markov chain as a stochastic process {SPn, n = 0,1,2, ... }, if Likelihood of{SPn4 i = j/SPn = i} = Likelihoodij where

(ij) indicates the movement between the location pairs over a given time interval from t to t+1. The continuous numerical dataset undergoes a process of partitioning the continuous variables to discretized intervals of equal lengths forming an interval valued partitioned vector. For the given dataset vector {di,d 2 ,...,dn}, determine the minimum, maximum and the range choosing the number of equal subintervals to be partitioned. Let us define and denote it as min{di,d2 ,...,dn}=a, max{di,d 2 ,...,dn}=b,

rang edi, d2, . . , dn } _ max~d 1 ,d 2 , . dn}-minadjd2 - adn } b-a, where 'si' denotes the number of subintervals. Si

number of subintervals. The interval valued partitioned vector formed will have finite set of points according to the number of subintervals chosen. For instance, if the number of subintervals is taken as 'si' , then the number of points in the interval valued partitioning vector will be 'si+1' and the union of all the subintervals must be equal to the original interval valued partitioned vector set and the intersection of these subintervals will be a null set. The interval valued partitioned vector set states are denoted by IVPV{ivpv 1 ,ivpv 2 ,...,ivpvS} such that ivpVr takes a value in the interval(aa+,)for1 r si,henceforminganinterval valued partition of the dataset. The midpoints Of IVPV{ipv 1 ,ipv 2 ,...,iVPVsi}is determined for symmetry of the balanced intervals. According to the number of movements of the location pairs (ij) over a given time interval from t to t+1, the Markov chain transition probability matrix 'MCTPM' is determined. The steady state probability matrix 'Pi' for the corresponding 'MCTPM' is calculated. This is the likelihood probability of long run. The midpoints of IVPV{ivpv 1 ,iVpV 2 ,...,iVPVsi} in interval valued partition vector set state having the highest probability is taken for prediction and analysis. Consider feature 1 in the dataset, calculate Markov chain transition probability of (n x n)dimension and steady state probability matrix in the long run of (1 x n) dimension for the feature 1 selected for prediction. Here, for feature 1, there are (nxn) unknown parameters to be optimized, considered as case 1. Now consider feature 2 in the dataset, calculate Markov chain transition probability of {(n - 1) x (n- 1)} dimension and steady state probability in the long run of [1 x (n -1)] dimension for the feature 2 selected for prediction. Here, for feature 2, there are {(n- 1)x(n- 1)} unknown parameters to be optimized, considered as case 2. Now the 2 matrices formed are with 2 different dimensions for the features considered and the two matrices are probabilistically related by their states.

A Rhotrix (3) is defined as a special matrix called coupled matrix 'cpldmtx' of dimension {[n + (n - 1)] x [n + (n - 1)]}, given two matrices of (n x n) dimension and {(n - 1) x (n - 1)} dimension respectively.

For instance, A Rhotrix R of dimension n can be written as

all a21 c11 a12

tt

1t

The element a (i,j =1,2,..., t) and ckl(k, =1,2,...,t- 1) are called major and minor entries of R respectively. Here for forming the Rhotrix, the two matrices are calculated by Markov Chain transition probability and then integrated with Rhotrix for predictive modelling. Calculate Markov chain integrated Rhotrix transition probability of {[n+(n- 1)] x [n+(n- 1)] dimension and steady state probability in the long run of {1 x [n+ (n 1)]}dimension for combined impact of feature 2 on feature 1 selected for prediction. Here, for combined impact of feature 2 on feature 1, there are {[n+(n- 1)] x[n+(n- 1)]} unknown parameters to be optimized, considered as case 3. All the three cases with unknown parameters to be optimized enter the Genetic Approach optimization procedure in the next phase for predictive modeling.

The drawings and the forgoing description give examples of embodiments. Those skilled in the art will appreciate that one or more of the described elements may well be combined into a single functional element. Alternatively, certain elements may be split into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, orders of processes described herein may be changed and are not limited to the manner described herein. Moreover, the actions of any flow diagram need not be implemented in the order shown; nor do all of the acts necessarily need to be performed. Also, those acts that are not dependent on other acts may be performed in parallel with the other acts. The scope of embodiments is by no means limited by these specific examples. Numerous variations, whether explicitly given in the specification or not, such as differences in structure, dimension, and use of material, are possible. The scope of embodiments is at least as broad as given by the following claims.

Benefits, other advantages, and solutions to problems have been described above with regard to specific embodiments. However, the benefits, advantages, solutions to problems, and any component(s) that may cause any benefit, advantage, or solution to occur or become more pronounced are not to be construed as a critical, required, or essential feature or component of any or all the claims.

Claims

WE CLAIM

1. A method for performing weather forecast, the method comprises:

partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector; initializing weather parameters in first phase using interval values Markov integrated Rhotrix predictive model; and optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

2. The method as claimed in claim 1, wherein the dataset vector determine minimum, maximum and range choosing number of equal subintervals to be partitioned.

3. The method as claimed in claim 1, wherein the interval valued partitioned vector formed is having finite set of points according to the number of subintervals chosen.

4. The method as claimed in claim 1, wherein modelling of Rhotrix predictive model comprises:

calculating two matrices by Markov Chain transition probability for forming the Rhotrix and then integrating it with Rhotrix for predictive modelling; calculating Markov chain integrated Rhotrix transition probability and steady state probability in the long run for combined impact of features on selected for prediction; and optimizing unknown parameters for combined impact of features and thereby entering genetic optimization procedure in the next phase for predictive modeling.

5. The method as claimed in claim 1, wherein steps for optimizing GA comprises:

loading input dataset thereby partitioning into equal sub intervals; calculating Markov chain transition probability of (nxn) dimension and steady state probability in the long run of (1xn) dimension for a first feature selected for prediction; calculating Markov chain transition probability of {(n-1)x(n-1)} dimension and steady state probability in the long run of [lx(n-1)] dimension for a second feature selected for prediction; calculating Markov chain integrated Rhotrix transition probability of{[n+(n-1)]x[n+(n-1)] dimension and steady state probability in the long run of {1x[n+(n-1)]} dimension for combined impact of the second feature on first feature selected for prediction; applying GA approach for parameter optimization of first case, second case, and third case as and thereby performing training and testing on the entire dataset using the Genetic approach operators and thereafter error measures are computed; comparing result obtained with the actual output using GA fitness function; and terminating optimization when stopping criteria, the average change in the fitness value less than options is reached and thereby displaying the predictive modeling output.

6. The method as claimed in claim 5, wherein there are (nxn) unknown parameters are optimized for the first feature, considered as first case, whereas there are {(n-1)x(n-1)} unknown parameters are optimized for the second feature, considered as second case.

7. The method as claimed in claim 5, wherein there are {[n+(n 1)]x[n+(n-1)] unknown parameters to be optimized for combined impact of second feature on first feature, which is considered as third case.

8. The method as claimed in claim 5, wherein the genetic operators used are selection, crossover and mutation operators.

9. A system for performing weather forecast, thesystemcomprises:

a pre-processing unit for partitioning continuous numerical dataset variables to discretized intervals of equal lengths forming an interval valued partitioned vector; a Markov integrated Rhotrix predictive model for initializing weather parameters interval values in first phase; and an optimization unit for optimizing estimation of weather parameters in second phase using genetic approach for reducing error in predictive modeling.

10. The method as claimed in claim 9, wherein the central processing unit is configured with the interval valued Markov integrated Rhotrix predictive model for initializing weather parameters.