Solution by Google colab

This assignment considers the following hypothetical scenario. A real estate agency would like to use artificial intelligence to better predict whether a certain customer will buy a specific house, so they can focus their efforts on promising potential sales. Specifically, they want to label pairs of customers and houses according to whether they belong to the target class buys or not. The agency has selected seven attributes, each taking values from { yes,no }, namely - Basic features of the house: garden: whether the house has a garden or not parking: whether the house has private parking or not good_nbhood: whether the house is in a good neighborhood or not expensive: whether the house is expensive or not - characteristics of the client: young: whether the client is young or not rich: whether the client is rich or not interested: whether the client is interested in the house or not They also have some preliminary ideas about the kind of model they are interested in and have collected a small dataset for machine learning. Your task is to help them understand their options better and to recommend next steps towards realizing their goal, by answering the following questions. Question 1 The agency is interested in Naive Bayes model using the three attributes expensive, interested and rich. Given the training data in Table 1 use your Knowledge on Narve Bayes classifier to predict the class for the data in testing data in Table 2. (1.2: 5 Marks) Based on a critical assessment of the Naive Bayes model considered above. Answer the following question a) Evaluate your Naive Bayes models from Questions 1 by calculating the accuracy on the test data provided in Table 2. (2.3: 1 Marks) b) Write a bullet point list of the advantages and disadvantages of the Narve Bayes classifier. (1.2: 2 Marks) c) Write the answer of each of the following questions from the agency: (1.2: 3 Marks) i. Is the Narve Bayes model a good choice to solve this problem? ii. If yes, why? iii. If no, why not, and what can we do to enhance this model? d) Do you have any further recommendations that could help us improve our use of machine learning for this problem? Demonstrate your answer. (2.2: 4 Marks) Table 1: Table 2: \begin{tabular}{|l|l|l|l|l|l|l|l|l|} \hline & garden & good nbhood & parking & expensive & interested & young & rich & buys \\ \hline 1 & yes & yes & no & yes & no & yes & no & no \\ \hline 2 & yes & no & no & yes & yes & yes & yes & yes \\ \hline 3 & yes & yes & no & yes & no & no & no & no \\ \hline 4 & no & yes & no & no & no & yes & yes & no \\ \hline 5 & yes & no & no & no & yes & yes & no & yes \\ \hline 6 & yes & yes & no & no & yes & yes & no & yes \\ \hline 7 & yes & no & yes & no & yes & no & no & yes \\ \hline 8 & yes & yes & no & yes & no & yes & no & no \\ \hline 9 & yes & no & yes & yes & yes & yes & yes & yes \\ \hline 10 & no & yes & no & no & yes & yes & no & yes \\ \hline \end{tabular}