Forecasting of industrial
indicators - a basis of increase of a production efficiency
Kamysbaev M. K.
The Kazakh national agrarian university
Of Almaty, Kazakhstan
In the conditions of a crisis situation in agrarian sector and
transition to market economy exclusively working out and realisation of
measures for stabilisation and maintenance of the further development with this
important for a society of sphere of manufacture has great value.
The basic
priorities of a modern agrarian policy of Kazakhstan is maintenance of food
safety of the country, formation of effective system of agrobusiness, increase
of competitiveness of a domestic production and escalating of sales volumes of
production both on internal, and on a foreign market, support agricultural commodity
producers.
In
these conditions executives and experts of agriculture by means of
economic-mathematical methods and models could define competently and
competently directions of development of branches of an agricultural production
taking into account requirements of the market, provide increase of
competitiveness, efficiency of an agricultural production and profit reception
by each agricultural enterprise.
Whether it is possible to expect, predict
approach of crises? Whether it is possible to be prepared for them or at all
them to avoid? Whether it is possible to reveal factors which define success of
economic development of the state or give to the enterprise, businessman
chances to grow rich? How to operate, to achieve well-being and success? To
answer these questions always it is possible: the success of any business is
half provided at the expense of effective forecasting and planning [1].
On the basis of the spent
analysis of a condition of methods of forecasting applied in republic
Kazakhstan and modelling of development of branches of agrarian and industrial
complex it is possible to draw following conclusions:
1. At the present stage
economic-mathematical models receive a wide circulation. However the
application condition in republic of methods of forecasting of development of
branches of agrarian and industrial complex leaves much to be desired. If
researches on forecasting at regional level at branch level such works
practically do not meet were spent and carried out. At forecasting of development
of branches of agrarian and industrial complex the approach to its realisation
should be sustained branch, instead of regional, especially at level of the
enterprises.
2. The basic direction was and
for the present there is a forecasting of development of agricultural machinery
and requirement for it, forecasting of productivity of agricultural crops and
some other agricultural objects.
3. In overwhelming majority
existing works are a basis of the mathematical decision of a problem in
agriculture only under one factor and one function, and calculations are
limited, as a rule, to a finding of numerical values of its parametres that
reduces their efficiency.
4. Told above in many respects
occurs in the absence of a technique and uncultivate methods of forecasting of
development of branches of agrarian and industrial complex. Hence, necessity of
working out of such technique and arises, at necessity, the software of this
technique.
The strict method of
forecasting on the basis of mathematical modelling and information technology
develops in our republic and the countries distant and the near abroad. Thus
researchers always prefer the analytical methods of mathematical modelling
allowing more widely to spend studying of various variants and more effectively
to operate by difficult complexes of the big system.
The considered approaches to forecasting give the basis to believe, that
the success of working out of the forecast depends on a correct choice of a
method of forecasting [2]. Attempts of some authors to give recommendations of
for choice corresponding method were unsuccessful, as were based on value
judgment of applicability of methods to this or that object of forecasting. Any
of them in itself is not suitable for all cases of prognosis experts. The
method choice depends, on our belief, from following factors: the purposes of
the forecast (task in view); time of anticipation of the forecast; specificity
of object of forecasting; reliability and completeness of the initial
information; restrictions of developers of the forecast (time of working out of
the forecast, algorithms, forecasting software).
It is obvious, that exact
coincidence of the fact sheet in the future and prognostical dot estimations is
improbable. Therefore the dot forecast should be accompanied by bilateral
borders - the interval forecast, - i.e. instructions of an interval of values
in which with a sufficient share of confidence it is possible to expect
occurrence of the predicted size. The increase in uncertainty of predicted
process with growth of the period of anticipation is shown in constant
expansion of a confidential interval.
In practical work the problem of quality of forecasts should be solved
is more often, when the anticipation period yet has not ended also actual value
of a predicted indicator it is not known. In this case more exact the model
giving narrower confidential intervals of the forecast is considered. In
practice not always it is possible to construct at once good enough model of
the forecasting, therefore the described stages of construction of prognosis
models of dynamics of an industrial indicator are carried out repeatedly.
Check of the general quality of the equation of regress is spent both on
value of factor of determination, and by Fisher's criterion. It is necessary to
notice, that t - the criterion of the Student is in a sense special case
F - Fisher's criterion. In these conditions, judging by corresponding
tables, the following parity takes place:
, 
The forecast of
indicators on animal industries has been calculated on trends to models of pair
regress with use of Statistical dialogue system STADIA, version 6.0 for Windows
which allows to choose approximating function from more than 20 functions.
At forecasting of indicators of animal industries on Kyzylordinsky area
by the technique resulted above following basic indicators have been
considered: livestock ÊÐÑ, including cows,
sheep and goats, horses, camels and birds for 1999 - 2008 (in all categories of
economy).
At forecasting of a number of
cattle on area the most adequate function - model logistical has been chosen:
Model: logistical 
coefficient. a b c d
Value 157,6 90,69 189,1-0,8363
The Item of fault. 1,192 2,648 61,26 0,05394
We mean. 9,198Å-6 2,211Å-5 0,02115 7,731Å-5
Source amount of squared. Grave.ñv Middle square.
Recourse. 1,062Å+4 3 3539,0
Residuum. 11,51 6 1,919
All 1,063Å+4 9
Êîððåë. îòíîø. h η2 F Îøèá. àïïðîê.
0,99946 0,99892 1844,0
0,43 %
Hypothesis 1: Regression
model is adequate to experimental data.
Almost functional value of the correlation relation and high value of
factor of determination h2 testifies to high
general quality of the constructed equation of regress.
Standard value of t-criterion at number of degrees of freedom 9 and a
significance value α = 0,01 makes 3,25,
and value of criterion of Fisher Ftable = 3,252 » 10,56. Hence, F> Ftable = 1844> 10,56. The
received t-criterion and F-criterion allow to assert about
probability 0,99, that in a general totality close correlation dependence takes
place.
Õýêñï Yýêñï Yðåãð the Rest Dover. èíò.
1 (1999)
157,9 158,7-0,8389 ±3,179
2 (2000)
160,0 160,1-0,1297 ±3,062
3 (2001) 164,6
163,2 1,4200 ±2,972
4 (2002)
170,2 169,5 0,7251 ±2,910
5 (2003)
178,6 181,0-2,3670 ±2,878
6 (2004)
199,5 197,9 1,5770 ±2,878
7 (2005)
216,2 216,4-0,2461 ±2,910
8 (2006)
230,8 231,1-0,2796 ±2,972
9 (2007)
240,1 240,0 0,1445 ±3,062
10 (2008)
244,5 244,5-0,004592 ±3,179
Xïðîãí Yïðîãí the Item îøèá Dover. èíò
11 (2009)
246,6 1,453 ±3,319
12 (2010)
247,6 1,523 ±3,480
13 (2011)
248,0 1,601 ±3,658
Alignment of a dynamic number
and the forecast of a number of cattle on area on logistical function are
resulted accordingly in drawings 1 and 2.
Drawing 1. Dynamics
of livestock ÊÐÑ on areas, thousand goals
(Logistical
function)
Drawing 2. The forecast of
livestock ÊÐÑ on areas, thousand goals
(Logistical function)
Calculations on other
indicators of animal industries have been similarly carried out. On animal
industries indicators it is necessary to notice that circumstance, that, unlike
plant growing, data give in to approximation by more simple functions that
allows to predict precisely enough on approximating functions both a livestock
of animals, and animal industries production. Hence, indicators on animal
industries are less dependent on inherent-climatic conditions, rather than
indicators on plant growing.
The literature
1. Bolshakov A.A., Karimov R.
N. Methods of processing of the multidimensional given and time numbers. - Ì: the Hot line - a Telecom, 2007. - 522 with.
2. Dubrov A.M., Mhitarjan V.
S, Troshin L.I.multidimensional statistical methods. - Ì: the Finance and statistics, 2000. - 352 with.
The resume
In article the analysis
of methods of forecasting and modeling of development of branches of
agriculture of Kazakhstan is carried out. Calculations on forecasting of
indicators of animal industries of Kyzylordinsk area are carried out.