Neural models in diagnostics of the financial result of housing and utility enterprises

The   aim  of  the  research  is  the  usage  of  an  artificial  neural network  as a tool not only for forecasting,  but also for operative diagnostics of a financial  state through combining  deterministic and  stochastic factors in one model.  This  circumstance  expands the possibilities of effective influence on the formation of an acceptable  level of the company’s  financial  condition  in various activities.  The  proposed universal  model  is presented  in the article in relation  to the  company’s  characteristics  in the  housing and  utilities  sector.

The  article  proposes a  method  for  diagnosing  the  level  of  the housing and  utility  company’s  financial  condition  based on the use of a factor neural model of the financial results of their activities.

Materials and methods.  The  neural network  modeling methodology allows you to create models that have  several advantages:  learning ability  (they  adapt  to various  changes);  universality  (able  to solve a wide  range of data analysis and processing tasks);  speed (process various  data  in  parallel  mode);  ease  of use (easy  to operate  after training);  fault  tolerance  (resistant  to local  damage  to the  neural network  structure and external  noise).

One of the main  tasks that neural networks  successfully solve is the problem of classification  – the assignment  of the sample  to one or several predefined classes. Most often, the input sample is determined by  the  input  data  vector.  The  components  of  this  vector  are  the various characteristics of the sample. The  classifier in the form of a neural network  relates the object to one of the classes in accordance with the partitioning of the N-dimensional input space. The  number of components of the vector determines  the dimension  of this space. In the context of this article, the input sample is the financial condition of the organization at a particular point in time. The input vector that characterizes the sample includes a set of direct and indirect factors of the financial results of a housing and utility company. The neurons of the  output  layer  are  a  set of different  classes.  In  the  course of operation,  the neural network  assigns to each input vector a neuron in the output layer. The significance of the input data can be adjusted using connections between neurons and changing the neural network architecture. Neural networks can have a complex architecture when different  parts of the  neural  network  include  different  numbers  of connections and different neurons.

The  article develops  the  ideas  laid  down  by its authors  in [7,  8], where a neural network  of direct propagation and a way of learning with a teacher have already been used. The model, described below, has been modified due to the authors’ desire to improve it, as well as dictated by the specifics of the housing and utility companies: a list of key indicators has been developed that affect not the financial  result, but, consequently,  the financial  condition of companies in this sector of the Russian  economy;  the number  of input factors characterizing the input sample was increased,  each direct factor or group of direct factors was supplemented  with an indirect factor; direct and indirect factors explaining the same processes are combined  into clusters that influence  the  corresponding  neuron;  the  number  of neurons  in  the output layer has been expanded, the number of classes has been increased, the data are classified by means of the neural network in more detail; in the course of the program, it is possible to select the period to which  the input data (month, quarter,  half year,  year) belong. The  additions  made  a positive  impact  on the  work  of the  neural network.  The  accuracy of attributing the input sample to a specific cluster and the sensitivity of the neural network  has increased. The number of clusters has grown up to 50. Innovations  have increased the  usability  of  the  program.  New   interface  allowed  to  analyze data  monthly.  The  programmatic  way  of interpreting the data  has changed  due to the fact that not all input data  changes depending on the period.

1. Anita Bai, Swati Hira, P.S. Deshpande. An Application of Factor Analysis in the Evaluation of Country Economic Rank. Procedia Computer Science. 2015. No. 54: 311-317. DOI:10.1016/j.procs.2015.06.036.

2. Barilenko V.I. Business analysis as a basis for managing sustainable corporate development. Menedzhment i biznes-administrirovaniye = Management and business administration. 2014; 4: 130-135. (In Russ.)

3. Vladimirova L.P. Prognozirovaniye i planirovaniye v usloviyakh rynka = Forecasting and planning in market conditions. Moscow: Dashkov and K; 2012. 308 p. (In Russ.)

4. Housing in Russia. 2016: Stat. col. Rosstat. Moscow; 2016. 63 p. (In Russ.)

5. Kallan R. Osnovnyye kontseptsii neyronnykh setey = Basic concepts of neural networks: tr. fr. Eng. Moscow: Williams Publishing House; 2001. 287 p. (In Russ.)

6. Kostina L.N., Gareyeva G.A. Neural networks in the tasks of forecasting time series. Innovatsionnaya nauka = Innovation science. 2015; 6: 70-73. (In Russ.)

7. Kurochkina I., Shuvalova E., Mamatova L., Kalinin I. Analysis model of the company’s financial performance based on neural network. 3rd International multidisciplinary scientific conference on social science & arts SGEM 2016, Conference Proceedings. Bulgaria: STEF 92 Technology Ltd.; 2016. Book 2. Volume 111. pp. 49-57. DOI:10.5593/sgemsocial2016B23

8. Kurochkina I., Shuvalova E., Mamatova L., Kalinin I. About diagnostics of an enterprise financial condition of the housing and public utilities. 5rd International multidisciplinary scientific conference on social science & arts SGEM 2018, Conference Proceedings. Bulgaria: STEF 92 Technology Ltd.; 2018. Volume 5. Issue 1.3: 3-10. DOI:10.5593/sgemsocial2018/1.3

9. Mak-Kallok U.S., Pitts V. Logicheskoye ischisleniye idey, otnosyashchikhsya k nervnoy aktivnosti. Avtomaty = Logic calculus of ideas related to nervous activity. Automatic. Ed. K.E K.E. Shannon and J. McCarthy. Moscow: Foreign Literature Publishing House, 1956: 363-384. (Translation of the English article). (In Russ.)

10. M. Tim Jones. Programmirovaniye iskusstvennogo intellekta v prilozheniyakh = Programming artificial intelligence in applications. Tr. fr. Eng. Osipov A.I. Moscow: DMK Press; 2011. 312 p. (In Russ.)

11. Nikiforova N.A., Dontsov E.V. The use of neural network modeling to predict the financial condition of the company. Upravlencheskiy uchet = Management Accounting. 2011; 4: 36-46. (In Russ.)

12. Osovskiy S. Neural networks for information processing. Tr. fr. Pol. I.D. Rudinskiy. Moscow: Finance and Statistics; 2004. 344 p. (In Russ.)

13. Plaskova N.S. Stochastic modeling and performance evaluation of a commercial organization. Upravlencheskiy uchet = Management accounting. 2014; 7: 59-67. (In Russ.)

14. Romanovskiy A.V. On the use of artificial neural networks to predict the financial performance of an enterprise. Audit i finansovyy analiz = Audit and financial analysis. 2013; 2: 363-370. (In Russ.)

15. Tadeusevich R., Borovik B., Gonchazh T., Lepper B. Elementary introduction to the technology of neural networks with examples of programs. Tr. fr. Pol. I.D. Rudinskiy. Moscow: Telecom Hotline; 2011. 408 p. (In Russ.)

16. Fedorov E.E. Iskusstvennyye neyronnyye seti = Artificial neural networks. Krasnoarmeysk: DVNZ «DonNTU»; 2016. 388 p. (In Russ.)

17. Fomin V.P., Fomin P.V. Analytical component of the successful management of financial results (conceptual aspect). Ekonomicheskiy analiz: teoriya i praktika = Economic analysis: theory and practice. 2015; 31: 13-25. (In Russ.)

18. Fomin V.P., Fomin P.V. Analytical component of the successful management of financial results (practical aspect). Ekonomicheskiy analiz: teoriya i praktika = Economic analysis: theory and practice. 2015; 36: 12-26. (In Russ.)

19. Khaykin Saymon Neyronnyye seti: polnyy kurs. 2-e izdaniye. Tr. fr. Eng. = Neural Networks: the full course. 2nd edition. Per. fr. Eng. Moscow: Williams Publishing House; 2006. 1104 p. (In Russ.)

20. Yakupov D.T., Rozhko O.N. Prospects for the use of artificial neural networks to predict the volume of freight in transport systems. Statistika i ekonomika = Statistics and Economics. 2017; 14; 5: 49-60. (In Russ.)