<b>Hey, the customs data you want is here! </b>

The force majeure of the customs data for several months has really worried the editor, and it has also brought a lot of inconvenience to the work of the bigwigs in the circle. In line with the principle of sharing concerns for customers and benefiting fans, the editor worked hard for several days, and made some model estimates on some customs data with a professional and objective attitude. In this issue, I will share with you a model of one of the products, and other product models will be reflected in the weekly and monthly reports produced by our company to solve the urgent needs of the big guys. the To be honest, forecasting is really not an easy task, so the editor cannot guarantee that the forecast is absolutely accurate, but it is absolutely reasonable. So please forgive me for the shortcomings. The following editor will take pure MDI imports from China as an example to tell you how to make deduction predictions. The overall process is roughly divided into three steps. The first step is qualitative analysis to determine the dependent variable (that is, the reference factor and its related data); the second step is quantitative analysis to make a fitting prediction of known customs data to verify the error range of the above dependent variable; the third step is quantitative Analyze and predict unknown import customs data to get the final result. 1.Qualitative analysis to determine the variable factors of reference the First of all, let's look at it qualitatively, to determine the variable factors that can be referred to with a large correlation coefficient, and first mine the imported data itself. the After dissecting the import data source of Pure MDI, we found that Japan and South Korea accounted for a very large proportion of the total domestic imports. Let’s feel the data picture wind.

China Import Customs Data (Korea)

China Import Customs Data (Japan)

China Import Customs Data (Japan & South Korea)

China's total import data

Proportion

201701

2151

3375

5526

6263

88%

201702

2870

4987

7857

8682

90%

201703

2928

4534

7462

8058

93%

201704

1802

4085

5887

6416

92%

201705

1345

2538

3883

4493

86%

201706

2215

1446

3661

4352

84%

201707

1736

2901

4637

5170

90%

201708

2606

4122

6728

7284

92%

201709

2483

4736

7219

7378

98%

201710

1158

3195

4353

4489

97%

201711

834

2821

3655

3796

96%

201712

1186

3646

4832

5013

96%

201801

2324

4509

6833

6975

98%

201802</sp; Second, after the variable factors are selected, we need to conduct quantitative analysis and estimate the future pure MDI import data. Here, the data processing tool of EXCEL is mainly used to make the theoretical model of linear regression. the The first step, let's do a verification first, the selected data sequence is as follows:

	South Korea & Japan Export Customs Data (China) (X)	China's total imports (Y)
201701	5525	6263
201702	7930	8682
201703	7112	8058
201704	5118	6416
201705	2816	4493
201706	3382	4352
201707	5029	5170
201708	6929	7284
201709	5939	7378
201710	4855	4489
201711	3397	3796
201712	5198	5013

Through EXCEL linear regression processing, the functional relationship between the variable factor "total exports of Japan and South Korea" and the dependent variable "total imports of China" is: Y=0.9531*X+927.51 the Then we will bring in the known data from January to March 2018 for verification, and get the following data, as shown in the table below:

	Exports to Japan and South Korea	Fitting Data	China's actual import	Deviation	Margin of Error
201801	6819	7426.6989	6975	451.6989	6%
201802	4707	5413.7517	5348	65.7517	1%
201803	6790	7399.059	6992	407.059	6%

It can be seen from the above table that the error between the predicted data and the actual data is 1-6%, which is in a relatively accurate range. Therefore, this method is used to estimate China's future in the next few months. pure MDI is more reasonable. 3.The final result of quantitative analysis and estimation of unknown import data the Repeat the operation of the first step above to expand the time span of the two sets of regression data to the existing March 2018. Next, we perform linear regression on the data from 2017 to the first three months of 2018 to obtain the functional relationship The formula is: Y=0.9175*X+1059.24 the Then, according to the total export volume of Japan and South Korea collected by us and brought into the above equation, it is estimated that China's pure MDI from April to June 2018 The total import volume is shown in the table below:

	Total exports to Japan and South Korea	China's Total Import Estimate
201804	5132	5768
201805	5111	5749
201806	5790	6372

In this way, the known data source - Japan and South Korea export customs data can be used to complete the unknown import volume of China's pure MDI for 3 months estimate. the Finally, to sum up, forecasting is risky, and you must be cautious when adopting it. The editor has no choice but to come up with such a move for you guys. The logic and rationality have been verified to a certain extent. There will definitely be discrepancies with the actual data, but it is absolutely reliable. Spectrum!s=link href="/MMDI/" target=_blank>Pure MDI Unknown 3-month import estimate. the Finally, to sum up, forecasting is risky, and you must be cautious when adopting it. The editor has no choice but to come up with such a move for you guys. The logic and rationality have been verified to a certain extent. There will definitely be discrepancies with the actual data, but it is absolutely reliable. Spectrum!

This article is from the Internet, does not represent the position of Epoxy curing agent, reproduced please specify the source.https://www.dmp-30.vip/archives/7916