Research Project Case Study Solution
This is one of the simplest forecasting methods. To calculate it, we add the demand for transformers from 2006 to 2008 and divide by 3 to forecast the data for 2009 to 2013. The analysis shows that the demand for transformers was 865 in 2009 and 899 in 2013. This method can be faster, simpler, and simpler to determine if the demand for a transformer decreases or increases depending on the mode. It is recorded by the moving average.
Descriptive Analysis:
Perform descriptive statistics to analyze the movement and nature of data between 2006 and 2010. The analysis shows that the average demand for transformers has increased over time, from 801,167 in 2006 to 928,574 in 2010. In the year 2006, the minimum demand for transformers was 695, while the maximum demand for transformers in 2010 was 1,153. The quality control department can use this data to analyze a company’s operational performance.
Exponential Smoothing Forecasting Technology:
Estimates recent inventory requirements using historical data. This method takes into account the most obvious observation factors and evaluates them accordingly. The formula used to forecast the transformer needed in 2011 is as follows;
Ft+1= αAt + (1-α)Ft
Whereas Ft is the actual demand of transformers and At is the foretasted demand of transformers. The analysis shows that in the financial year 2011, the maximum demand for transformers was 1129. Exponential smoothing provides a more accurate estimate as it takes into account the forecast and actual values for the most recent period. Then. In short, this method takes advantage of the difference between the expected actual value and the actual value.
One sample T- Test:
One sample T- Test is used to analyze whether the average value of the transformer increases. The analysis shows that the critical value of t is greater than t, the value of p is 3.8%. Compared to the 5% significance level, this value is not high, which means that it provides strong and important statistical evidence to reject the t value. The null hypothesis indicates that the average number of transformers did not change significantly. Similarly, it indicates that the number of transformers may be equal to or greater than 1000 transformer units.
One-way analysis of variance:
This is a widely used assumption-based statistical technique. It is used to assess whether the number of transformers has changed from 2006 to 2010. The test shows that the p-value is 0.010, which is below the 5% significance level, which means that it provides a strong and important statistical basis. We reject the null hypothesis that the average number of transformers did not increase between 2006 and 2010.
Regression Analysis:
This is the most effective statistical tool for predictive modeling that can help examine the relationship between more than two interesting variables. Investigate the relationship and causal relationship between the dependent variable (demand for regulators and televisions) and the independent variable (sales of refrigerators). The analysis shows that the correlation between the sale of refrigerators or televisions and the demand for voltage stabilizers is 92.5%, which is positive, and therefore shows that the sale of refrigerators directly affects the number of transformers required. In addition, the adjusted square of R is 85%, which means that the percentage change in the demand for the voltage regulator is high based on sales of TVs or refrigerators. The regression coefficient measures the extent to which a change in one variable predicts a change in another variable. The regression coefficient is 0.314, which means that if sales of frozen TVs increase by one, demand for voltage stabilizers will increase by 0.314 (Schleifer, 2017).
Caesars Entertainment
To find the casino registration as a dependent variable, the independent variables are the type of customer and the average daily fee the customer paid for the room, because in the type of casino, knowing the type of customer is very important. From this, the casino will find out which type of customers visit the casino most often. This knowledge will help you determine the casino’s target market. In addition, the average daily fee paid by the guest for the room will help determine the relationship between the daily fee and check-in, thereby helping to determine the appropriate daily fee, thereby increasing the casino’s registration rate. Increase your income and profits.
The null hypotheses for these independent variables are
- There is no relationship between the type of the customers and the casinos check ins.
- There is no relationship between the average daily rate paid for the room by customers and the casinos check ins.
- The average check ins in all days (Monday to Sunday) are equal.
The alternate hypotheses for these independent variables are
- There is a relationship between the type of the customers and the casinos check ins.
- There is a relationship between the average daily rate paid for the room by customers and the casinos check ins.
- The average check ins in all days (Monday to Sunday) are not equal.
The customer's casino type coefficient (total reward card holder) is 0.37, which means that if 1 customer logs in to the casino and has a customer called a casino, this will result in the total casino check in. The amount has increased by 0.37 checks.
The customer's SE (Special Access Event) type coefficient is 0.03, which means that if 1 customer registers in the casino and has a customer type called SE, the total casino registration will increase by 0. 03 record.
The coefficient for segment 0-3 of customer 0 (in customer groups 0, 1, 2 or 3) is -0.07, which means that if 1 customer logs in to the casino and the customer type is segment 0-3, reduce the total casino registration with 0.07 registration.
The customer’s FIT (Free and Independent Traveler) coefficient is 0.22, which shows that if 1 customer registers at the casino and has a customer named FIT, the total casino registration will increase by 0.22. -tendon.
The customer (group member) group type coefficient is 0.2, which means that if 1 customer registers in the casino and has a customer named Group, the total casino registration will increase by 0.2 registrations.
The customer’s WS (wholesale customer) coefficient is 0.02, which means that if 1 customer registers at the casino and has a customer named WS, it will increase the total number of casino registrations by 0.02 records.
The ADR FIT coefficient (average daily fee paid by free and independent travelers for a room) is -0.52, which means that if the average daily fee paid by free and independent travelers in a casino for a room is $ 1, the casino the number of tickets is increasing. Entry decreased by 0.52 logins.
In addition to this, there is no need to worry about variables such as the type of guest and the average daily fee the guest pays for the room, or its definition and use in the regression model.
In addition to this, the one-way ANOVA has been performed in the excel spreadsheet and found the p value is 0.0000 which provide the statistically strong evidence to accept the alternate hypothesis which states that the average check ins in all days (Monday to Sunday) are not equal and reject the null hypothesis which states that the average check ins in all days (Monday to Sunday) are equal.
Fueling Sales at EuroPet
The regression analysis between the sales and fuel volume shows that the p-value is lower than the significance level of 0.05 which provides the strong statistical evidence of rejecting the null hypothesis which states that there is no significant relationship between sales and fuel volume. Additionally, the t-test is below than 2 which provides strong evidence that there is a significant relationship between fuel volume and sales.
The 95 percent confidence interval for the three estimates are calculated with the use of the formula provided below;
95% Confidence interval = mean (X) +/- t s/SQRn
For 95 percent confidence interval, the z-value is 1.96. The confidence interval of sales at the minimum, maximum and average fuel volume level is 63306 and -62398. The point estimate for the true mean fuel volume level in the sample is 63306 and we are 95 percent confidence that that the true mean is between 63306 and -62398.
In addition to this, it has been found that there is a strong relationship between fuel volume and sales as the sales volume is influencing the c-store sales. Additionally, the c-store sales would be increase of €0.65 for each additional fuel liter sold. Furthermore, a fuel volume explicates the 37 percent variation in the sales at c-store, but is not enough to claim the casual relationship between variables.
Furthermore, the regression analysis is performed to assess the relationship between sales and two advertising variables such as radio and TV. The regression analysis shows that the p-value is below than the significance level of 0.05 which provides strong statistical evidence of rejecting the null hypothesis which states that there is no relationship between advertising variables and sales whereas the t-value of TV is greater than 2 which means that there is a significant effect of TV on sales. On the other hand, the t-value of radio is less than 2 which means that there is an insignificant effect of radio on sales.
The regression equation is as follows;
Regression model = -3745.2+(5.59*TV)+(0.27*fuel volume)+(85.17*fuel price)+(68.93*temp)+(-138.5*prec)+(895.25*holiday)+(-157.25*visit)
There would be an increase in sales of €5.59 for every increased add. Additionally, there would be an increase in sales of €68.94 for every increased degree. There would be an increase in sales of €0.28 for every liter sold. Similarly, for every €1 in price of fuel, there would be an increase in sales of €85.18.
The unintuitive sings include holidays which shows that there would be an increase in sales of €895.2 if it is holiday. Similarly, every add mm of precipitation would lead to the reduction in sales of €138.50. Every added person comes ½ times per week would lead to the reduction in sales of €157.2.
Conclusion
Considering that A-Cat is developing a more accurate model for forecasting transformer demand, it is recommended that the company use exponential smoothing forecasting technology as it is easy to apply and can generate forecasts. always accurate. The forecasts in the series can be expanded to support data with seasonal components. Similarly, this forecasting technique allows company management to plan inventory levels more efficiently and relevantly with the latest data.
By taking into account two data, the actual value of the last period and the forecast value of the similar period, it also reduces the need for data storage. The use of a low smoothing constant means that the forecast will be less sensitive and therefore means a low peak demand for the transformer. If the product generates a lot of noise, the company can use a smaller constant. On the other hand, if the noise level of the product is low, the company may apply a higher constant. Effective use and application of this technology allows the company to make accurate future forecasts for the product (John C. Chambers, 2015).
According to Europet’s analysis, it has been found that the variables that have a significant impact on c-store sales include temperature, fuel prices, TV ads and number of visits. Therefore, the company is recommended to increase its efforts only in TV commercials. As a result of television advertising, the brand will increase or predict its visibility, thereby increasing in-store sales c. Because trial variables can manipulate results, companies should not use these variables.
After calculating the MAE (Mean Absolute Error), MAPE (Mean Absolute Percentage Error), Maximum Error and Minimum Error, it has been found that the MAE and MAPE for moving average model is lower than the linear regression model. Liner regression model has 380 MAE which is higher than the moving average model’s 190 MAE. In addition to this, the MAPE for linear regression model is 46 percent which is also higher than the moving average model’s 21 percent MAPE (Vandeput, 2019).
The maximum and minimum errors of the moving average model are also smaller than the maximum and minimum errors of the linear regression model. The maximum error of the moving average model is 771, which is less than the maximum error of 819 of the linear regression models. In addition, the minimum error of the moving average model is 0, which is smaller than the minimum error of the linear regression model 5.
These analyzes help determine that the moving average prediction model is better than the casino’s linear regression prediction model.
Although the moving average forecasting model is the best time series forecasting model for casinos in a given situation, based on the shortcomings mentioned at the top of the document, the moving average model responds slowly to rapid price changes. which often occur at market turning points. To avoid this, casinos need to use a more advanced time series prediction model, which is an exponential moving average prediction model.
The advantage of the exponential moving average forecasting model is that it reacts faster to price changes than the moving average forecasting model, taking into account recent price changes. This is especially useful for traders who try to trade at higher and lower levels during intraday fluctuations, as the exponential moving average pattern changes faster than the moving average pattern...............
This is just a sample partial case solution. Please place the order on the website to order your own originally done case solution.
Related Case Solutions & Analyses:
ATH MicroTechnologies Inc. (B)
New Leaders of Financial Giants: The Cases of Vikram Pandit (Citi) and John Thain (Merrill Lynch)
How to Perform Sensitivity Analysis with a Data Table
History of Investment Banking
The Decision to Denuclearize: How Ukraine Became a Non-Nuclear-Weapons State
Kidnapping Negotiation (B)
Disentangling the Web: Losing Control and Loving It
Bank of America-Merrill Lynch
VICTORIA HEAVY EQUIPMENT COMPANY
Merck Sharp & Dohme Argentina, Inc. (A)
