Анализ базы данных

Question 1A (10 points) - there are 4 "business related" questions.Considering the IBM_HR_Attrition business case and data (do not open Excel when answering the question - you may use MongoDB and Dataiku only) 1. (2.5 points) What did you choose to be your "business success" variable (choose something different than Attrition)? Can you please define it? Why did you choose that variable? 2. (2.5 points) Provide a count of possible unique business outcomes? How many business successesdoes your data have? 3. (2.5 points) Is there any interesting business insight when you calculate the average commute per your "business success" buckets? (In other words ... what is the average commute for your business success and business failure and what is the insight? )4. (2.5 points) This question will depend on the environment you chose:-If you chose Dataiku answer this question only: Why did you NOT choose MongoDB to find answers to these questions?-If you chose MongoDB answer this question only: Why did you NOT choose Dataiku to find answers to these questions?(Word count: 50-75 for each question) Question 1B (10 points)Please provide any code you have designed for Question 1A.  Explain your code.  Question 2A (10 points)Using MongoDB Compass (or alternatively the online Mongo Atlas):1. How many Pokemon (how many documents) are Flying or Grass (in Type_1) and have Attack greater than 65?  What insight does it bring? 2. What is the average Speed of all the Pokemon that have Attack greater than 75 and are not Legendary? Explain how you can interpret the results.   You will need to provide the MongoQL syntax in the next question. PLEASE SAVE YOUR CODE!(Word count: 75-150 for each question) Question 2B (10 points)Please paste any code that you have designed for Question 2A and explain which sub-question does it belong to.  You will receive 5 points for correct code for Question 2A.1 and another 5 points for correct code for Question 2A.2  Question 3A (10 points)Consider your Pokemon data in Hadoop (in Dataiku). Answer the following questions using your PySpark templates! Alternatively, you can build your model in the Dataiku LAB. You should build a logistic regression model to predict the value of Legendary = 1, by using Sp_Attack. Do NOT split your data into training and testing!! Use all the data to train your model! Dependent variable (Y) : "legendary"Independent variable / feature (X1): "sp_attack"  In you answer, please provide the model output, including all the coefficients and AUC ROC for this model. (You will receive max points for accurate results.)   Question 3B (10 points)1.(5 points) Can you explain any business insight from the model? what is the interpretation of the coefficient of sp_attack? 2. (5 points) Provide the code that you've used to build your regression in Question 3A. (copy and paste the code below your business insight).

вот так выглядят вопросы на которые мне нужна помощь с ответамитак же есть файлы которые я так понимаю и есть база данных которую нужно анализировать через Dataiku или MongoDB