Environmental Science & Classification Decision Trees Lab

Environmental Science & Classification Decision Trees Lab

Environmental Science & Classification Decision Trees Lab

ORDER NOW FOR CUSTOMIZED AND ORIGINAL ESSAY PAPERS 

The questions will require you to have knowledge and carry out various features of R for data analysis, including data wrangling, feature engineering, resampling methods, GLM, GAM, MARS, decision trees and neural networks. Of course, you have to implement the correct techniques that is relevant to the question so you have to clearly understand to fluently answer the questions. I am having trouble with these questions as I haven’t been up to date with the lectures in this module due to my thesis deadline.

IMPORTANT: Think through all the steps in data analysis and explain why you are doing what you are doing. Provide me with explanatory text, R code and the outputs from R to support your answers. Above all, answer the question posed! The relevant excel sheets required for the questions will be attached in this request. Please do not place an offer if you don’t understand the questions properly.

5attachments

Slide 1 of 5

• attachment_1attachment_1
• attachment_2attachment_2
• attachment_3attachment_3
• attachment_4attachment_4

QUESTION 1 [50 MARKS]                                        

The standard way to assess the age of the Scarlet Clam (a shellfish) is to count its growth rings. This method is accurate but slow and tedious.

You have been given a contract to devise a new method. The client believes it should be possible to build a predictive model for age based on various body measurements and has gathered together the data for you (available on Blackboard).

The Slobodkin Institute has provided data on adult male and female clams in two separate csv files (Shellfish_males.csv and Shellfish_females.csv). These have the same format with columns as follows:

A third file has been provided by Cobalt Marine and gives data for infant clams which cannot be sexed as they are too young. Cobalt Marine have assembled the data in an Excel file (Shellfish_infants.xlsx) but with a different format (variable names are supplied).

Combine all the data and build a predictive model for age. You may assume that your model will be used by a competent researcher. You should design your model to need as few measurements as possible while also achieving good predictive power.

In reporting your findings back to the client, you must justify your choice of modelling technique, assess the importance of each predictor variable you have included, and state the likely overall predictive power of your model when applied to new data. Fundamentally, you must answer the question whether counts of growth rings can be replaced by a model based on body measurements.

QUESTION 2 [50 MARKS]

You have been employed by a city council to advise them on their transport policy. Specifically, they want to know whether certain types of vehicle should be banned from the city centre on the basis of their CO₂ emissions.

As CO₂ emissions data are not available for every vehicle, they have asked you to find other, more general evidence that would provide criteria on which vehicles could be banned.

To help you to do this, they have supplied you with a dataset (Carbon_emissions_21.xlsx) showing the CO₂ emissions for more than 39,000 vehicles manufactured from 1984 to 2019. This data set contains information on the CO₂ emissions and other characteristics of the vehicles.

Carry out any analysis that you consider appropriate to identify the characteristics of the most polluting vehicles (in terms of CO₂ emissions). You must justify your choice of modelling technique and provide evidence on how successful your model is.

Use your findings to provide a set of guidelines to the city council to identify the characteristics that could be used as a basis to ban certain types of vehicle when the actual CO₂ emissions are unknown.