I need my data science assignment solved. It is about supervised learning with regression and classification.

Please find my assignment attached. It has to be done in Python.


Get Help With a similar task to - I need my data science assignment solved. It is about supervised learning with regression and classification.

Login to view and/or buy answers.. or post an answer
Additional Instructions:

Assignment 2 EM-0212-02-Applied Data Science Due: March 11, 11:59pm For this assignment you would be working with the wine dataset. The idea is apply supervised learning techniques to get some useful inferences from the dataset. The dataset is available directly from sklearn (datasets.load wine()). Question 1: Regression For this problem just consider the 13 attributes (ignoring the class). The idea is to predict the Alcohol value using other 12 attributes. So, in essence you have 12 features and 1 target. By providing evidence from your analysis, answer the following questions (you can either code the regression model yourself or use sklearn), 1. Are all features providing useful information ? 1 2. In your opinion, which feature is most helpful in determining the alcohol value ? 3. Which feature is least helpful ? 4. If we consider a polynomial model, what degree polynomial would you recommend ? 5. Instead of polynomials if we use trigonometric functions, will it help ? 6. In your opinion, what information, the following observations will give regarding the alcohol content (higher or lower) • Increased value of Color intensity • Reduced value of Proline • Increased Magnesium but reduced Ash value Question 2: Classification Now considering all 13 attributes as features and class as the target, fit a K- Nearest Neighbor classification model to answer the following questions 1. In your opinion, what would be a suitable training and testing split for this dataset and why ? 2. Do you need all 13 features to get the best performance out of your model ? 3. If you can get similar performance with less number of features, which model would you recommend (using all the features or using less features) and why? 4. Are any samples wrongly classified in the test dataset ? If yes, write a code to get the correct labels and the reason (according to you) behind this wrong classification. 5. Using the proportion of samples correctly classified (in the testing data) as a measure of performance, provide a plot showing the relationship of performance vs K (number of nearest neighbors) 6. Do you have any recommendations for improving the classification model performance? 2

Related Questions

Similar orders to I need my data science assignment solved. It is about supervised learning with regression and classification.
20
Views
0
Answers
Excel Index and match remove
Remove extra text besides part number in () and then match or index item to OH inv1 and inv2 on sheet...
60
Views
0
Answers
data analysis project for Intro to Data Science
Answer all the questions according to the instructions....
26
Views
0
Answers
Visual Studio "Creating your first app" w/reveal SDK
I am attempting to create https://help.revealbi.io/en/developer/web-sdk/create-first-app.html#step-1---create-the-project. I have a dashboard completed and the code is listed as follows but I am dealing with a lot of errors and bugs when adding the sugges...
39
Views
0
Answers
homework assignment mapping data in r studio
need the whole homework assignment done in an .r file and in a knitted html file showing the graphs and explaining briefly, and answers....