Assessment 3-Take-home- 30%
Please provide your answer for each question. For the coding questions, you are only allowed to use R programming. You also need to include your code in the answer sheet.
Question 1 (10)
What is the prescriptive analytics and what is its role in business analytics? What are main methods in prescriptive analytics (describe about each method with some examples).
Question 2 (10)
What is the stochastic optimisation? Provide an example of real life considering the stochastic optimisation.
Question 3 (10)
What does simulation important? What are the applications of simulation in your daily life? Provide an example.
Question 4 (10)
Provide an application of Expert systems in your daily life. What are the features of such system and what are the challenges of using such system (advantages and disadvantages)
Question 5 (15)
We assume that a person wants to determine whether or not they should go on a hiking trip (that H is true). The person will consider the following factors: weather (W), available budget (B), travel possibility (T), and college degree (C). Here is the rules:
R1: IF the available budget is more than $1000 and his age is between 18 and 50 then it is possible to travel.
R2: IF the person could travel and the weather is good then the person should go on hiking trip.
R3: IF the person has a college degree then the available budget is more than $1000 R4: IF the age in between 18 and 30 then the person has a college degree.
Imagine the person is 20 years-old and the weather is good. Using backward chaining to find out if the person should go on hiking trip or not.
Crating the rules using the defined letters. Provide the graphical inferencing.
Question 6 (15)
Define a real-life problem in any domain and use non-linear optimisation to solve it.
You need to specify an objective function and constraints based on the problem you have found. Use R for solving the problem. Please avoid using the newspaper project for answering this question.
Question 7 (15)
A hospital wants to determine the optimal staffing levels for its nurses and physicians in order to minimise costs while maintaining a certain level of patient care. The hospital has identified two types of nurses, RNs and LPNs, and two types of physicians, MDs and DOs. Each RN can handle up to 5 patients per day and each LPN can handle up to 3 patients per day. Each MD can handle up to 20 patients per day and each DO can handle up to 10 patients per day. The hospital wants to ensure that each patient is seen by a physician and a nurse at least once per day. The hospital's objective is to minimise the total cost of staffing while maintaining a high level of patient care. The number of patients each day is approximately 1000 patients.
The hourly wage for RNs is $25, the hourly wage for LPNs is $20, the hourly wage for MDs is $100, and the hourly wage for DOs is $80. The hospital has a budget of $50,000 per day for staffing.
Please provide the objective function and constraints in R, as well as the R code and solution. Assume that each person can work 8 hours a day.
Question 8-(15)
Suppose you are working as a data scientist for a large e-commerce company that generates a massive amount of transaction data every day. You are tasked with analysing this data to gain insights that can help the company improve its business performance.
1- What are some of the challenges you might face when working with big data? How can you address these challenges?
2- Suppose the company's transaction data consists of 1 billion records per day, each record containing 100 fields. What is the total size of the data per day? What are some of the storage and processing requirements for analysing such a large amount of data?
3- Suppose the company wants to perform real-time analysis of the transaction data to detect fraud and other anomalies. What are some of the challenges of real-time data processing? What are some of the tools and techniques you might use to address these challenges?
Note: You can assume that the necessary computing infrastructure is already in place to handle the large volume of data.