TW201216182A - A prediction system and method for trend of fashion colors - Google Patents

Abstract

A system and method for predicting the trend of fashion colors is developed to determine the upward or downturn of the trend for a specific color in the future. Firstly, the related historic data regarding the developing trend for the color are processed and generated as a series of parameters. Then a quantitative methodology theory-based on gray theory is employed to forecast its trend during a specific time of the future by using the obtained parameters.

Description

201216182 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a prediction system and method for popular color trends, which predicts and applies impact factor analysis for color pop-up influence factors. Φ [Prior technology]

In general, the preference behavior data of the popular color market is often irregular and belongs to random changes. For the process of random variables, most of them use the method of probability statistics to analyze and study, trying to find out the law, but the results are not good, mainly because the method of probability statistics is an abnormal distribution and non-stationary process for the analysis of popular color preference distribution analysis. It is very tricky and has been difficult to overcome. For example, the Bayes' theorem method must be considered when the probability of pre-existence and the conditional probability are known. However, the random variables affecting the popular color, such as social culture and political and economic situations, will affect people's changes in color psychological needs and preferences, but these Random variables with random variables are difficult to define, and it is more difficult to find some relevant probability values beforehand. Therefore, using the Bayes' theorem method in statistical methods has some limitations in practice; The premise of feasibility and accuracy is that there must be a large amount of raw data, and the innate restrictive factors for the non-regular and inability to widely collect information on popular colors. This is also the main reason why academics and enterprises are still unable to construct a system for effectively predicting popular colors. 5 [S] 201216182 At that time, for the trend forecast of popular color, commonly used quantitative basic methods, including regression analysis (including linear regression, multi-factor regression, single-factor regression, stepwise factor regression, non-stepwise factor regression), analysis of variance, etc., regression Although it is a more general method, most of it is only used for small factors, linear, and difficult to deal with multi-factor, nonlinear systems, and the popular color market is full of uncertainty and non-existence as described above. Linear system.

It is generally believed that there are many shortcomings in using the regression analysis of the trend market. First, it is necessary to predict the trend of the trend to have a large amount of data. If the amount of data is small, it is difficult to find the statistical law. Furthermore, the regression analysis requires that the distribution of the data must be linear, exponential, or logarithmic, because the linear regression is easier to calculate, and most people want the distribution to be linear, if the regression is single factor or less. The analysis may also be index or logarithmic distribution, but overall, the requirements of regression analysis are more typical, not chaotic. In addition, the amount of calculation in the regression analysis is large, and the calculation of the regression analysis of single factor or two factors is not too large. But when the factors are more than two, the calculation work will be huge enough to be done manually. It must be assisted by a computer. In view of the above, we use regression analysis to predict the market for popular color prediction. There are indeed many difficulties. But can the market's popular color trends be predicted? Whether the prediction results are accurate or not has been different for a long time. The pattern of predicting and analyzing popular colors has always been the subject of research and solutions by different industry-related companies, and this is the problem that the present invention intends to improve. SUMMARY OF THE INVENTION The object of the present invention is to provide a gray prediction module for predicting 6 [s] 201216182 sub-analysis to import the output value of the data, and the actual implementation of each module is imported into the original Applicable (()> system and method for predictive application of influence factors for color prevalence change influence factors. The prediction system and method for color fashion trends of the present invention first color related data of past trend development trends, and then A plurality of parameter values are generated for the operation of the relevant line characteristic parameters, and finally a gray prediction module is established by using the parameter data, and the gray prediction module operation can predict the trend of the color in a certain period of time in the future. The present invention is not limited by the following detailed description of the embodiments of the present invention, and is not intended to limit the present invention. It is one of the examples of the prediction system of the color trend in the present invention, mainly for a specific single color. Measure, it contains the application as follows: 1, start the prediction module 100 to start detection.

2. The relevant data module of the predicted color development trend over the years 20 0, the user wants to predict the color of the development trend of the color over the years X (0> · 3, the screening module 300, the relevant raw data Filtering, screening and making necessary adjustments, and then generating several parameter values for the calculation of the relevant raw data X characteristic parameters. 4. Gray prediction module (Gray Model) 400, using the data of the parameter values to establish a Gray prediction module (Gray Model), and by using the gray prediction 201216182 module operation output, a predicted value 1(1) of a generated Λ value predicting the trend of the color in a certain period of time. Λ 5, storage media database The module 500 can derive a final predicted value X by using the predicted value I(1) of the generated value, and then the final predicted value is stored in the storage media database module to provide query and analysis. The prediction module 600 is terminated to end the detection.

The invention discloses a method for predicting a color trend, which comprises the following steps: 1. First, the step 1 0 0 starts the prediction method. 2. Perform step 2 0 0 to import the relevant raw data X(Q) of the trend of the user's desired color over the calendar year. 3. Then proceed to step 300, filter the relevant raw data, screen out the inapplicable ones and make necessary adjustments, and then generate several parameter values for the relevant raw data X (Q performs characteristic parameter operations. 4. Then Step 400 is performed to establish a gray prediction module (G ray Μ ode 1) by using the data of the parameter values, and calculate the output by the gray prediction module to predict a generated value of the color trend in a certain period of time. Predicted value / (1) Λ 5. The predicted value I(1) of the generated value obtained in step 400 can be used to derive a final predicted value /, and then step 00 is stored in the database of the storage medium. To provide query and analysis use. 6. Finally, step 60 is used to end the prediction method. 201216182 FIG. 2 is a schematic block diagram of the modeling process of the gray prediction module of the present invention, and the gray prediction module is clearly known. The modeling process of 0 firstly generates the generated value X(l) from the original data X(G), and establishes a computational model for the X(l) using the gray prediction 模组 module to obtain a prediction of the generated value of the model. Value I(1) and one a The a and b coefficient values of the model equation, and then the predicted value I(1) of the generated value is reduced Λ The predicted value I(Q) of the original data is compared with the original data value X(G) to obtain a residual value. ΔΧ(()>, the residual value ΔΧ(()) is applied to test the accuracy of the gray prediction module, and φ uses the coefficient values of a and b of the module equation to determine the positive or negative trend of the future color trend. The intensity, and a predicted value X is obtained from the predicted value JT(C) of the original data. The gray prediction module 400 treats the stochastic process as a change in a certain range, a gray quantity process related to time, Only the original data X (G> has more than 4 pens, the generation of the gray prediction module can be established by using the generation conversion. The invention uses the gray prediction module as the analysis method in the prediction system, and changes the influence factor of the color trend. Analysis of impact factors.

3 is a schematic diagram of a storage medium database module of a preferred embodiment of the present invention, a storage medium 700, wherein a computer program 7 1 0 and a database 7 2 0 are stored for loading into a computer system 8 0 0, and perform a color trend prediction method and store the final predicted value /. The invention is also a prediction system for color popularization and best-selling trend, and a method for using the same, the relevant original data X (G> of the color is related to popular and popular color survey statistics published by popular institutions and magazines, etc. The data β 9 [S] 201216182 empirically found that the use of gray prediction modules can achieve quite good results in color trends. This system and method is feasible and can obtain good prediction results. The gray prediction module is based on random originals. The time series, the new time series representation law formed by time accumulation can be approximated by the first-order linear differential equation solution. It is proved that the original time series revealed by the first-order linear differential equation solution exponentially changes. The prediction of the grey prediction module will be extremely successful when the original time series implies an exponential change law. In summary, the present invention is a prediction system φ for color trend and a method of using the system, the final prediction The value X is the original data x (G> is predicted by the gray prediction module operation to predict the future Value (ie, popular color or best-selling color survey statistics) and the a and b coefficient values of the constructed predictive model (determining the trend up or down and its intensity). The overall forecasting method is not due to differences in personal emotions or preferences. It affects the correctness of color trend forecasting. The gray forecasting module can integrate, verify and adjust according to the change of objective environment and the best-selling color of the market for actual consumer acceptance, and predict the color of the same period in the next year. The most likely trend of coloring completely excludes personal emotional factors and market interference, and is more capable of integrating the advantages of φ with theory and practice, and the operation is also simpler, so it can achieve the purpose of the present invention. The preferred equivalents of the invention are not limited thereto, and the simple equivalent changes and modifications made in the scope of the invention and the description of the invention are still covered by the invention. [Simplified illustration] 10 [S] 201216182 FIG. 1 is a preferred embodiment of a prediction system of the present invention Schematic diagram of a flow chart for a color trend; FIG. 2 is a schematic block diagram of a modeling process of a gray prediction module in a preferred embodiment of the prediction system of the present invention; FIG. 3 is a gray diagram of a preferred embodiment of the prediction system of the present invention. Schematic diagram of a storage media library module of the prediction module. • [Main component symbol description] 100 ..... Step 200 ..... Step 300 ..... Step 400 ..... Step 5 0 0 .....Step 600 ..... Step

700 .....storage media 710 .....computer program 720 .....database 80 0 ..... computer system: raw data 1 (1): generated value

II

[S] 201216182 / (.) : Predicted value of raw data /(1) : Predicted value of generated value ΔΧ(0 >: residual value X : final predicted value 3 · coefficient value b : coefficient value

12

[SI

Claims (1)

201216182 X. Patent application scope: 1 _ A forecasting system for popular color trends, including the following modules: a start-up prediction module; a data module for predicting the development trend of color over the years; a screening module; a gray prediction module Group (Gray Model): A storage media library module. 2 _ In accordance with the color trend prediction system described in claim 1 of the patent application, wherein the start prediction module is used to initiate detection. 3. According to the color trend prediction system described in item 1 of the patent application scope, the relevant data module of the predicted color development trend of the forecast is related to the development trend of the color that the user wants to predict over the years. Raw data X(0). 4. According to the prediction system of the color trend described in item 1 of the patent application scope, wherein the 'screening module 3 0' filters the relevant original data, screens out the inapplicable ones and makes necessary adjustments' The raw data is subjected to a characteristic parameter operation to generate a plurality of parameter values. 5. A prediction system for color trends according to the scope of the patent application scope, wherein the gray prediction module (Gray M〇del) 'creates a gray prediction module using data of the parameter values (Gray M〇de丨), and by the gray prediction module operation output, a predicted value %(1) of the generated value predicting the trend of the color in a certain period of time. 6. The color trend prediction system according to the scope of the patent application scope, wherein the storage media database module uses the predicted value of the generated value χΛ(1) to derive a final prediction value, and then The final predicted value is stored in the storage media library module to provide query and analysis usage. 7. A method for predicting the trend of popular colors, including the use of a prediction system for color trends as in claim 1 of the patent application, and performing the following steps: (a) importing data on the trends of the color over various time periods in the past; (b) generating a number of parameter values by performing a characteristic parameter operation on the relevant data; and (c) Using the parameter values to calculate a predicted value that predicts the color's trend over a period of time. 8. According to the method for predicting the trend of popular color according to item 7 of the patent application scope, further comprising a step (d), which compares a predicted value of the plurality of pens with the original data, in addition to evaluating the accuracy thereof. In addition, you can use it to adjust the final forecast to increase accuracy. 9. According to the method for predicting the color trend described in item 1 of the patent application scope, the predicted value obtained in step (c) can be stored in a database for providing query and analysis. 1 0. A storage medium comprising a computer program and a database for loading into a computer system and performing a prediction method of a popular color trend as described in claim 7 and storing a predicted value . 14 [S]