Decision Analysis for the United States Army - Essay Example

? Decision Analysis for the United s Army Case analysis for the U.S. Army Decision analysis tools have been applied for inventory control, resource allocation, scheduling of funds, replacements amongst other problems. The army has not been left behind in applying decision analysis tools for problem solving. Uncertainties and unpredictable situations are experienced in the army since war cannot be forecasted on the expenditure it incurs. Weapons acquisition is among the problematic areas in the US department of defense that has decision analysis tools applied. Analytical decisions are undertaken when there is ample time to compare the facets, in order to choose the most profitable course of direction. The best module amongst the alternatives is taken (See appendix). In this case the supply of weapons is done from the highest bidder and of the best quality. Decision trees This is a tool that allocates data for decision making in form of tree in order to analyze the variable hence make the decision. The tool breaks down the data in logical and simple manner thus easy to understand. Any mistakes made in the data translate to the tree model. Each nodule of the tree represents an attribute, the branches corresponds to an attribute while each leaf assigns a classification. The module is very simple to use and interpret and it can handle both numerical and categorical data. The decision trees can be used for validation analysis as other statistical approach follows (Andrew, 2005). The simple illustration below describes the case of buying or making weapons. A decision is made after thoroughly considering all the uncertainties involved then a decision is made. Cost factor can be used to make the decision and the manufacturing option would be adopted. Nonetheless, other factors need to be considered in real situation analysis e.g. quality, time taken etc. Illustration one: A variable is an element of a problem that is being predicted or determines it. They may change depending on the scope of the problem being analysed. Predictable variables are used to assess the certain conditions while unpredictable variables follow uncertainties hence probability is used. Variables are used in decision analysis to represent value or symbol of the actual attribute being determined. A variable denoted by Y may represent the cost estimates, overheads, revenues etc. When one value of a variable is related to another value of a variable, then the case is said to a correlation between the two variables. Correlation means an inter-relationship. A simple linear regression model is a statistical technique that explores the inter-relations between two or more variables. Linear regression tools have applied in many organisations to forecast for the observed variables. The military may want to determine whether there is a relationship between warring period and weapons use. To do so, the observed data is analysed and the simple line equation is used. Y=a+bx Y in this case represents number weapons used during war X variable shows the number the war reoccurs A and b are constants Table one weapons Time period lasted during the Wars 100000 1 year World war I 300000 1 year World war II 250000 2 years’ War in Afghanistan 85000 1 year War in Kuwait The equation is then extend to determine a and b i.e. 100,000= a+ 1b 250,000=a+2b -150,000= -b 150,000= b Then a=100,000-1b = 100000-150,000 = -50,000 If we were to determine the occurrence of world war III will take 3 years, then the number of weapons needed are: Y=a+bx = -50,000 + 150,000*3 =400,000 weapons The simple regression analysis can be determined by plotting a graph using the same variable to obtain the line of best fit. This can be then be used to forecast on future needs of weapon acquisition for the military. The relationship between the period the war takes the higher the number of weapons needed. The relationship is linear. This model works best when there are two variables under consideration. Scatter diagrams are used to analyze the relationship between two variables. The variables are plotted in the graph and the line of best fit cuts through the majority of the variables. As the case of simple regression analysis, the scatter diagrams are used for predicting variables and relationship between variables is also assessed. Multiple regression analysis and its uses Multiple regression models are applied where there exists more than two variables under consideration. In this case the function changes to: Y=a + b1X1 +b2X2 In the example above, if an extra variable of number of lives lost is added here then the solution would be different in determine the constant variables as shown below: weapons Time period lasted during the Wars Lives lost 100000 1 year World war I 1 million 300000 1 year World war II 2 Million 250000 2 years’ War in Afghanistan 3 million 85000 1 year War in Kuwait 0.8million The layout to be followed is to obtain X1Y, X2Y, X1X2, ?X1, ?X2, ?Y. To predict the unknown variables the equation below is used: bx1= (n?x1y-?x1?y)/{n?x21 –(?x1)2} The equation is complicated thus computer softwares have been developed to determine the constants a and b. The model allows the management to analyze the behaviour of each individual model towards the yield (David et al, 2007; Christian et al, 2008). Inventory control allows managers to determine the lead time i.e. the time between the supply of raw materials and production time. The lower the lead time the more effective the company operates because it is will always be on time in production and meeting the consumers demand (Donald, 2003). Just in time module is applied when the production is made on delivery of raw materials. In this case inventory is not held in the stores. An ABC tool helps an organisation to categorize its resources according to the order of importance. These two tools assist the management of a company to control resources-one of the vital tool being time for production. Enterprise resource planning is beneficial since it individualizes the needs of an organisation. It assists the management to budget effectively and adopt the most viable decisions in running the organisation (Susan and Belverd, 2010). To forecast means to predict the future outcomes. This is achieved through the use of qualitative and quantitative methods. Quantitative forecasts are based on historical data to predict the future. Time series forecast follows observation and survey of various outcomes. Qualitative forecasts have a limitation of data hence judgment and experience is used in this case e.g. introduction of a new product in the market. Exponential smoothing follows an automatic weighting of past data with weights that decrease exponentially with time i.e. current values greater weights. This is best used when they exists seasonality of factors, rise and fall of demand etc. Delphi method is follows a procedure where panel of experts answers questions independently. It is beneficial in that the best decision can be made and answers can be shared to further discuss the problem at hand (Terry, 2006). Appendix References Andrew L., 2005, Decision analysis: an integrated approach,4th Ed., John Wiley and Sons, Chichester. 23 Christian A., Wayne, L. & Christopher, Z., 2008, Data Analysis and Decision Making with Microsoft Excel 3rd Ed., Cengage Learning, London. 596 David, G., Lawrence, L. & Keith, E., 2007, Applied regression analysis and other multivariable methods,4th Ed., Cengage Learning, London. 25 Donald, J., 2003, Inventory control and management2nd Ed., John Wiley and Sons, Chichester. 7-15 Susan V. & Belverd E., 2010, Managerial Accounting, 9th Ed., Cengage Learning, London. 184-186 Terry, L., 2006, Quantitative techniques, 8th Ed., Thomson Learning, London. 169-171 Read More