Purpose: Introduction/background of problem
As a consultant for Excellent Consulting Group, we were tasked with developing and
testing different forecasting techniques. In case study number three we utilized linear regression
as a form of forecasting. The information used to test this assessment was gathered by our client
who collected data on sales of one of its products, a lottery app for smart phones and hits on its
website. This information was gathered over a 12-month period and includes the hits and sales
for the corresponding months. The consulting manager at Excellent Consulting Group has
requested we go the next step and analyze additional forecasting methods. As consultants we
feel it is important to do a
…show more content…
The same data will be utilized in this case study but simple exponential smoothing will be utilized to help analyze the data. The information we decided to focus on for the purpose of this case was the sales data from our client`s lottery app. We will use this information to calculate the mean absolute percentage error (MAPE) for the first twelve months assuming the calculation for January represents its actual sales. Two different alphas 0.15 and 0.90 are utilized in order help determine our assessment.
In the below spread sheet the formatted data are to the left and the data are arranged by months
and sales. Immediately to the right of this information is the excel chart utilizing alpha 0.15 to
forecast and analysis the data. To the right of that is similar information but the alpha utilized in
this scenario is 0.90. The analysis with the lowest MAPE will help us determine which
forecasting equation achieves the best outcome for our analysis. Looking at the diagram it
appears that utilizing 0.15 is a better forecasting method than utilizing 0.9. As you can tell, mean error (ME) were almost four times higher using the alpha 0.9. The
average sales minus forecast sales using alpha 0.15 totaled -33.404. The mean percentage error
(MPE) average utilizing alpha 0.15 was -0.105 and -0.061 utilizing alpha 0.9. The MAPE
| B. Yes, because the test value 1.257 is less than the critical value 1.782
o In summary this analysis shows the percent of every dollar in sales that is
in a business. For example if another organisation sold 1000 sandwiches in a week they would decide to order more sandwiches for the next week. However this data needs to be accurate and reliable, for example if the data they received told them they sold 0 sandwiches they wouldn’t order any more
Analyzing the data that clarify the points of strength and weakness not only for of the organization but the competitors as well. this mean for each piece of information collected for the organization, the team concern by the data collection has to obtain the revaluate data for the compotators. "A critical evaluation of the organizations past performance, present condition and the desired future conditions must be done by the organization. This critical evaluation identifies the degree of gap that persists between the actual reality and the long-term aspirations of the organization." [3]
The average of the trials is 14.1 m/sec2. Assuming there could be about a 3% error estimate above the actual value, the actual value is probably between 13.7 m/sec2 and 14.5 m/sec2.
(b) Consider the percentage error obtained in Part B. What may be responsible for the difference in the values between Part A and Part B?
The objective of this experiment was to find the best possible model for forecasting. I will use a series of tests both visual and statistical to find the absolute best model for forecasting the data set. The forecast will be made for the conglomerate Wal-Mart. I start my test by taking the time series plot graph of the data. This indicates whether the data has a seasonal or quarterly trend, and if there is a time trend. I also run a trend analysis on the data set. I compare the graphs through trend analysis, and choose the graph with the smallest amount of error. I have elected to use the quadratic trend model because the mean square deviation (MSD) is much lower in terms of error compared to the linear graph. My other objective, is to
We shall identify statistical tools and methods to collect data and also analyze data to determine the appropriate decision for the identified problem in A-CAT CORP forecasting.
It is characterized as efficient utilization of accessible data to recognize dangers and to gauge the danger [ISo/IEC Guide 51:1999,
The data may be placed into a line graph to look at the way that each month compares. This is shown in table 2.
Neural networks provide an alternative solution to the traditionally used statistical methods of forecasting. Traditional method models include variances of linear
In this method three different participants take part first participants are those that are decision makers and they make the final decision, Second participants are staff personal assist these people collect the surveys result and the number of the possible questions the third participants are respondents they provides the valid information to the decision makers before the forecasting decision made people (J.S Armstrong, 2001).
The article posted by http://www.entrepreneur.com/article/41384. The title of the article is “How to Start a Consulting Business” Obviously like most people I was reading one day and happened across this article. This article had a lot of information and really had me turning my gears. Here is a bit of information that compelled me to want to do more and to do this business feasibility plan. In 1997 U.S. businesses spent just over $12 billion on consulting. According to Anna Flowers, spokesperson for the Association of Professional Consultants in Irvine, California, the association has recently noticed an increase in calls for information from people who want to get into the business.
This prediction would be based on the assumptions followed by the acronym LINE. Linearity, Independence of errors, Normality of error, and Equal variance. The first assumption of linearity is indeed linear. As the number of hours of telemarketing increases the number of new subscription also increases as seen in the graph attached. By plotting the residuals in order of sequence in which the data was collected this would complete the check of independence of errors. As seen in the residual chart attached there are no relationships between consecutive results. Normality is evaluated by tallying the residuals into a frequency distribution. The residual data in the frequency distribution attached shows that they follow a normal distribution. The final assumption of Equal variance shows that there are no major differences in the variability of the residuals. Therefore, by passing all four tests of
This is an assessment which indicates how well a company did or the organization to achieve what the customer’s needs and expectation. This indicates if the service that a company or organization is good enough to impress their customer or must be improve for the better.