In current work problems and requirements for demand forecasting in commercial or manufacturing enterprises are analyzed and suitable forecasting algorithms are proposed. In enterprises with multidimensional and heterogeneous demand it is advisable to use different algorithms for different demand constituents and to readjust parameters used for forecasting. Existing forecasting packages are not practical as they are not integrated with commodities or materials supply orders management activities and business processes of enterprise. The orders management system is developed with forecasting component using adopted time series forecasting techniques such as moving average, exponential smoothing, double exponential smoothing, seasonal etc. These techniques ensure reliable forecasting results for different time series models: random, trend, seasonal and cycling, and are integrated with other business management activities. It is possible to calculate deviations of forecasted demand from factual values, to select algorithms giving minimal perсentage error, and to adjust algorithms parameters to changing demand. The system can help managers to choose forecasting algorithms and to adapt their parameters in the course of time. The system is designed using UML CASE tool and implemented in Microsoft .Net environment using MS SQL Server for data storage.

Vienas iš darbo tikslų yra apžvelgti rizikos sąvokos aiškinimus, jos rūšis, veiksnius bei vertinimo ir analizės metodus. Šio darbo tyrimo tikslas – vertinti pasirinktos įmonės X riziką pagal finansinių ataskaitų duomenis. Siekiant įvertinti įmonės riziką ateičiai, darbe prognozuosime 2007 -2008 metams statistinius įmonės dydžius: nuosavo kapitalo vertę, grynąjį pelną, įsipareigojimus, turtą ir kt. Pagal realius ir prognozuotus duomenis skaičiuosime finansinius rodiklius. Prognozes atliksime pagal tiesinius ir netiesinius prognozavimo metodus. / Risk is interpreted and explained very various. The risk is necessary in Lithuanian economical system, because the produce are begun to make earlier than demand begin to grow. The purpose of this work is to review interpretation of risk, it’s kinds, factors, methods of analysis and estimation. In this work, risk of the firm estimate by statistical data of financial accounts. We forecast statistical firm’s data: the capital, the net profit, the liabilities, the assets and etc.; for 2007 and 2008 years. We compute financial indexes by real and forecast data. In this work, we forecast number by trends with seasonal decompositions, curve of Regression, Moving Average method, Simple Exponential smoothing method, Autoregressive model and Autoregressive Integrated Moving Average (ARIMA) model. We use the statistical and analytical software package STATISTICA. After forecasting number of assets, liabilities we got that Polynomial trend with seasonal components is the best model for them. Logarithmical trend with seasonal components is the most infallible model for capital’s forecasting number. The best model for interest is Autoregressive model. The best model for net profit is Regression model. Risk of the firm is estimated by calculated financial indexes ROCE, ROA, ROI.