Business Proposal Editing

Attached are th data and question, Please help me with soulution using R program langauge Serving up Data: Application of data in the Hospitality Sector __________________________________________________________________________________ Soumya Roy and Soumyadeep Kundu wrote this exercise solely to provide material for class discussion. The authors do not intend to illustrate either effective or ineffective handling of a managerial situation. __________________________________________________________________________________ As Sudhir stepped out of the conference room, he realised he had quite a task ahead. In a week’s time he would be making a presentation to one of the largest hotel chains in India. The chain was looking to better utilise the years of Property Management System (PMS) data from its hotels to make better business decisions and for more efficient operations. They had approached Sudhir’s firm, a leading analytics and consulting concern, to assist them and provide some use cases of the data. Sudhir, a senior associate, had been tasked with preparing the pitch. During the meeting, Sudhir’s boss had made it abundantly clear the importance of the pitch during the upcoming meeting with the chain, since they would be signing on as a client only if they were convinced that the firm could help them leverage their data. For Sudhir this was a challenge having never worked on any assignments involving the hospitality sector. To add to the woes, the chain had declined from sharing any data from its PMS. Sudhir would have to dig deep to prepare the pitch and develop a credible application of the data. The Indian Hospitality Industry The hospitality industry occupies an important place in the Indian service sector accounting for 6.8% of the national GDP and 8% of all employment in the country1 . The industry had been growing rapidly as well, fuelled by both domestic and international tourism. With the advent of aggregator services such as OYO and AirBnB, the market had opened up new categories. Large hotel chains such Ginger Hotels, ITC Hotels, Taj Hotels, JW Marriott too operated across the country. Being such a burgeoning sector, the Indian and various state governments have a number of programs to attract tourists and also propagate the industry such as Incredible India, Swadesh and others. Though the recent pandemic had taken a heavy toll on the sector, it was gradually reverting back to its pre-pandemic levels with people venturing out again. Analytics in the Hospitality Industry The hospitality industry has recently opened up to the use of analytics for more efficient operations and better decision making. Numerous hotel chains such as Marriott, Hyatt make use of their PMS data for revenue management. The data from the system such as room preferences, cancellations can help hotels better manage revenue. The data is also used for targeted marketing campaigns, better personalised services, targeted discounts and offers. All of these uses culminate in better guest experience, which is important in an industry which relies on positive word-of-mouth and garnering favourable reviews online2 . 1 Tourism & Hospitality Industry of India, Indian Brand Equity Foundation, [login to view URL] 2 Tripadvisor Reviews and Online Reputation Management, Carreiarao Paula, [login to view URL] The Challenge As Sudhir dug deeper into the hospitality industry and its many nuances, he realised there were a number of opportunities for the use of PMS data. In his research he had encountered the pestering issue of hotels having to deal with cancellations. Rooms being perishable commodities if not sold for a night would not generate any revenue. For hoteliers, maintaining profitability involved maintaining a strong ADR (Average Daily Rate) which was the average revenue earned from an occupied room per day, and Occupancy Rate, which is the percentage of occupied rooms at the hotel in a given time. The revenue for each room would be ADR times Occupancy Rate. Therefore in the event of a booking being cancelled or no-shows, both the ADR and Occupancy Rate would be affected and in turn the revenue. Worldwide, cancellations was a major problem with cancellation rates being as high as 40% in 20193 largely owing to zero to minimal cancellation charge policies. In India too, cancellations pose a problem to hoteliers who often don’t overbook. Sudhir now had a use case for the presentation. Fortunately for him he had found a dataset4 . Though the dataset was from the Portugal, it contained the requisite fields (Exhibit-1). As Sudhir began work on developing the models, he believed he had the perfect demonstration using the data he had discovered. Exhibit 1: Data Description Variable Type Existing/New Definition ADR Numeric Existing Average Daily Rate, which is the average rental revenue earned for an occupied room per day. Scaled between 0 and 1. Agent Categorical Transformed Whether Booking is done through an Travel Agent or not Children Categorical Transformed Whether guests are bringing children Company Categorical Transformed Whether booking was made through any company/entity and billed. DOW Categorical New Day of the Week Days In Waiting List Categorical Transformed Whether the booking was in the waiting list before it was confirmed to the customer Repeated Guest Categorical Existing Value indicating whether the booking was done by a repeated guest or not Lead Time Integer Existing Number of days that elapsed between the entering date of the booking into the PMS and Arrival Date Room Type Categorical New Whether the Reserved Room Type and Assigned Room Type are same or not Cancellation Status Categorical Existing Whether the booking is cancelled or Not Source: Authors 3 Cancellation rate at 40% as OTAs push free charge policy, Hertzfeld, Ester. [login to view URL] 4 Antonio, N., de Almeida, A. and Nunes, L., 2019. Hotel booking demand datasets. Data in brief, 22, pp.41-49. Assignment Questions 1. Build a basic logistic regression model using only the "Agent" feature (booking with or without agent) as a predictor for cancellation status (cancelled or not). Extract the resulting logistic regression equation from the fitted model. Use the fitted model to calculate the probability of cancellation for two scenarios: a. Booking made through an agent. b. Booking made without an agent. 2. Build a logistic regression model using the predictor “Lead Time” to predict the cancellation status. Provide the fitted logistic regression equation. Interpret the results. 3. Discuss in detail the model building strategy for predicting the cancellation status for each hotel booking. 4. Divide the data set into two parts --- a training set consisting of 80% observations and a test set consisting of remaining 20% observations. Fit a logistic regression model using the relevant predictors based on the training set. Examine the model output and interpret the coefficients related to significant predictors in detail. 5. Judge the performance of the model based on the test data set, after identifying the optimal threshold point. Is the performance of the model satisfactory? Discuss in detail. 6. Provide the confusion matrix for the test data set. Calculate the sensitivity, specificity, and total error rate for the test data set. Comment on the performance of the model. 7. Discuss in detail how the chosen model can be used in making managerial decisions in this context.