I need to forecast Total Agricultural Tractors sales Volumes

[dariobalmas]

Guys,
I need your suggestion about how to go on.
I have 10 years of Total Agricultural Tractors sales Volumes. I have just 10 numbers (full years)
I need to forecast this year.
I thinking to find 10 years variables in internet like (GDP, Commodity prices, Wheater info, Subsidies and so on.
The variables could be:

• Numerical ones like commodity price, GDP
• Nominal ones like Weather (dryness, …), Subsidies (YES or NO)

Then based on this year variables forecast I want to find this year Total sales volumes.

How I can do? Regression?

[mlauber71]

@dariobalmas, welcome to the forum.

There could be several ways to do this. We recently had two discussions about regression predictions with historical data. You could check them out and use the accompanying workflows as an example:

Unsupervised machine learning (gradient boosting regression)

I set up an example that should look like your data (if not feel free to adapt and inform). We have two years of data (from one city) and data from January and want to predict the coefficient for Feb 2020. [image] 2019 serves as training and 2018 as test data. You might have more complex settings with cross-validation if you want. And also since you have the data you might provide coefficients of some more past months. Each line represents its own time series so to speak. Of course, such a se…

mlauber71/Public – kn_forum_predict_sales_coeff

Can KNIME be used to show the employment and unemployment rate in the UK from the year 2008-2018, showing the prediction for the next year?

I set up an environment that first transforms you data from 2008-2016 into a table that has the Area (region) as an ID since I understand that is the unit you would want to predict. [image] The Target is taken from 2017 “Emp UK percent”. If you want Unemployment you would have to change that. The years are transformed into columns, while the first year of the data becomes _0 and the year before that _1 and so on. [image] Then I construct a Test dataset using the data of past 9 years from …

mlauber71/Public – kn_forum_unemployment_uk

[dariobalmas]

I am try to understand the correlation betweet Target and many variables like this:
Can you help me?

Year	Target	GDP (current USBiolion$)	Agriculture, forestry, and fishing, value added (% of GDP)	Exports of goods and services (% of GDP)	Imports of goods and services (% of GDP)	Cereal production (metric tons)	Crop production index (2004-2006 = 100)	Cereal yield (kg per hectare)	Food production index (2004-2006 = 100)	Livestock production index (2004-2006 = 100)	Land under cereal production (hectares)	Subsidies and other transfers (current LCU)	Wheat	Crude Oil	Beef
1994	2299	139.8	4.214955475	21.47418652	19.29476677	15985688	90.11	2585	82.96	77.89	6184004	57.5	150	46	2
1995	3405	155.5	3.537162821	22.13559863	21.47534006	7514387	73.02	1419.2	70.91	73.38	5294961	10.2	177	45	2
1996	4574	147.6	3.860002427	24.08026334	22.58706515	13670070	95.24	2490.2	85.26	74.71	5489640	102.8	208	31	2
1997	2034	152.6	3.692557569	23.9873307	22.85793118	13254451	94.76	2271.8	85.88	76.55	5834396	107.3	159	43	2
1998	2196	137.8	3.448135515	25.00505476	23.89156288	10232323	88.95	2181.7	80.66	75.05	4690000	117.7	126	70	2
1999	1880	136.6	3.237963805	24.69520924	22.16667685	10059495	95.59	2191.4	86.95	80.28	4590484	123.4	112	62	2
2000	1940	136.4	2.994745986	27.15888005	24.27889391	14549009	102.68	2765.9	95.47	87.93	5260171	136.0	114	62	2
2001	2567	121.5	3.216383213	29.37482794	25.42680549	10714748	93.32	2424.1	88.71	84.74	4420161	150.9	127	67	2
2002	1498	115.5	3.384390416	31.78083655	27.98379975	13048225	102.11	2772.6	96.45	87.66	4706204	174.0	148	83	2
2003	6400	175.3	3.052348605	26.88506116	24.51676993	11819153	99.74	2536.8	96.9	91.23	4659074	201.6	146	117	2
2004	4765	228.6	2.763467723	25.46709621	25.61093793	12026899	100.31	2777.9	98.19	94.33	4329495	234.8	157	142	3
2005	3438	257.8	2.389623573	26.44666415	26.70245124	14175144	106	3309.6	102.85	99.57	4283072	266.8	152	150	3
2006	3193	271.6	2.331412417	29.27389366	31.00337066	9452300	93.69	3141	98.96	106.11	3009317	317.6	192	198	3
2007	6062	299.4	2.643372701	31.17385051	32.50923678	9507863	92.88	2790.3	101.74	112.56	3407418	367.9	255	129	3
2008	5938	286.8	2.859484952	35.62243807	37.24295217	15339478	112.73	4063.2	118.24	122.02	3775240	443.7	326	168	3
2009	2767	295.9	2.713427607	27.91188817	27.50637343	14570877	109.03	4405.5	116.5	123.05	3307450	507.2	224	214	3
2010	3709	375.3	2.387372914	28.61523338	27.37375243	14700931	107.73	4150.4	117.58	127.21	3542031	543.2	224	222	3
2011	3328	416.4	2.285472538	30.46094059	29.65168931	12928365	107.49	4017	116.13	125.24	3218442	618.6	316	222	4
2012	7620	396.3	2.169477737	29.72387936	31.17582008	14556168	112.4	4243.5	119.47	127.76	3430267	676.1	313	205	4
2013	6170	366.8	2.097756311	30.97134112	33.2704187	14154568	116.64	4043	123.17	129.26	3500995	756.9	312	118	4
2014	1919	350.9	2.174706761	31.46864521	32.9658559	16619889	123.42	4896.2	126.7	129.49	3394444	806.5	285	100	5
2015	2275	317.6	2.088751184	30.15288833	31.46418515	11917922	113.43	3531.5	122.05	131.92	3374717	869.4	204	124	5
2016	3938	296.4	2.22405748	30.580928	30.05726005	10187273	104.81	3818.6	116.66	131.1	2667828	908.1	167	157	4
2017	5768	349.6	2.361236373	29.62766937	28.34622556	18905871	…	5648.2	…	…	3347247	979.6	174	144	4
2018	5946	368.3	2.176019651	29.9070826	29.56325111	…	…	…	…	…	…	…	210		4
2019	3802	…	…	…	…	…	…	…	…	…	…	…	202		5

[mlauber71]

Question is do you want a correlation or a prediction. For correlation there is a node that would give you a matrix, prediction you could follow the last example.

Linear Correlation

You can see which variables have the strongest correlation with your Target, which would be Wheat, GDP with positive and “Agriculture, forestry, and fishing, value added (% of GDP)” with negative correlation (the larger your target the lower this value).

Linear correlation – mlauber71

[dariobalmas]

Great
I would like to predict 2020 value target based on p-value/correlation index of variable like GDP.
I have already 2020 GDP (and other variables…) forecast
Does is make sense?
Thanks
D.

[mlauber71]

@dariobalmas

Does is make sense?

Well you could try to do that. In this example I have used the data from 2015 and 2016 as test and the rest as training. With the model and the data for 2018 you could also predict the target. You just have to be careful what your target means. If it is the sales from 2018 the question would be when would you know the Target and when the other data. So in oder to make an actual prediction you would need the explaining variables before you would need the prediction (Target).

I the example workflow the best model is H2O GLM. Although the prediction for 2015 is not very good.

Row0 is 2015 and Row1 is 2016 (H2O omitted the RowIDs). 2016 is OK, 2015 not so much.

And as has been mentioned in the other discussions I referenced. With such few data points and a complicated market it is not so easy to make an accurate prediction. Some additional data might help. Or you could eliminate non-standard years (if there is such a thing).

compare several regression models on historical "Total Agricultural Tractors...