The Computer Programming Language Cheap Assignment Help
Please email Us your Assignment email@example.com
Call Us: +44-7497786317, +61-280147200
What you have to work on is the only bonus question. You have the ability to deal with SPSS, Excel, and interpret them correctly. If you got wrong, I ask a refund. Before starting it, if you cannot do, just switch to the other writer who can deal with SPSS and Excel. In order to solve the bonus question below, you have to look at critically following resources below.
2) Abstract for the research paper
3) the excel for the question to be solved
4) the answer from1 to 4 for the question ,which other did.
5. Calculate “probability of being active” over time for each household. Please report
A. How you calculated the probability
You only do B and C correctly 100%. If it is incorrect, I will ask 100% refund by showing the answer later. This is only two questions based on what I uploaded. Do draw graph for them.
B. A time series graph of randomly selected five households
C. A time series graph of mean probability across households.
Select households that purchase the product category at least two times in the observation period. Convert the dataset to apply the BG/NBD model proposed by Fader, Hardie, and Lee (2005, Marketing Science). Using SPSS is recommended, but you may use other programs as well. Please explain how you selected the households and converted the dataset. Please also report the first 10 rows of the converted data. You may assume that the first purchase of each household in the observation indicates the starting point of transactions data for each household.
In the beginning, I chose the sample duration for data that gathered between 200601 and 200653 duration (52.14 weeks, 1years).
I divided the data based on each customer and encapsulated them.
The number of purchase is expressed as x. t_x is last transaction of customer (in 2006 years), and T is (52.14 – day of the first purchased).
By implementing the computer programming language (php), I took them to .xls file
Estimate four parameters of BG/NBD model using Excel Solver as guided in Hardie’s notes. Report the parameters with the maximized log-likelihood function value
Real purchase data:3287 transactions
Repeat transactions are purchased by more than twice visits users. Therefore, there are differences between real purchase data and repeat purchase data.
Randomly select 5 households from the dataset (converted in Q1). Calculate conditional purchase expectations of these households between the week of 200814 and 200834 based on information available up to the week of 200813. How accurate is your expectation?
Forecasting Purchase of random customers
Random samples are not exactly accurate when predicting the real transactions.
Let’s look at Column3 which is model versus Column4 which is real data.