SIT 720 Machine Learning Assessment

Part 1: Linear Regression: (25 marks)

1. Load the dataset and split the data for training and testing - consider the data of last 2 years (2015 and 2016) for testing. Now exclude recording_date_time column from both training and test sets. Display the shape of training and test sets. (3 marks)
   In [ ]:
   # INSERT your code (or comment) here
2. Consider the 'temperature' as the target. List the insignificant features for predicting temperature, if any. Explain your findings. (5 marks)
   [Hint for students: See the "7.3 Relevance and Covariance among features or variables" for more information.]
   In [ ]:
   # INSERT your code (or comment) here
3. Now create a linear model considering the 'temperature' as the target variable and other columns as features (you can optionally remove non-contributing features). Show the test performance (as Mean Absolute Error, MAE) of the model. (5 marks)
   In [ ]:
   # INSERT your code (or comment) here
4. Find the feature which shows maximum correlation with "pressure". Create a linear regression model to predict temperature using these two features ('pressure' and the one which shows maximum correlation). Compare the performance of this simplified model with the model developed in the previous question (Q-3). Explain the performance variation, if any. (6 marks)
   In [ ]:
   # INSERT your code (or comment) here
5. Apportion the complete dataset into training and test sets, with an 40-60 split. (6 marks)
   1. Train a linear regression model without considering overfitting scenario and report the test performance.
   2. Create an optimal regularized linear regression model and report the test performance.
   3. Explain the reason behind the performance variation, if any.
   In [ ]:
   # INSERT your answer in maximum five sentences.

Part 2: Logistic Regression: (9 marks)

1. Can the same target (temperature, mentioned in Part-1) be used for logistic regression?
   Why? (2 marks)
   In [ ]:
   # INSERT your code (or comment) here
2. Split the dataset as 70-30% for training and testing. Create a logistic regression model to predict the 'precip_type'. Report the prediction accuracy of your model whether the "precip_type" is "rain" or not (use decision threshold of 0.45). (5 marks)
   In [ ]:
   # INSERT your code (or comment) here
3. Discuss the test performance using precision, recall and confusion matrix. (2 marks)
   In [ ]:
   # INSERT your code (or comment) here

Part 3: Objective function optimization: (6 marks)

Let’s consider the line graphs shown below and answer the following questions [Hint: See weekly content 7.4-7.10]

1. Which of the above figures represents the convex objective function and why? (1 marks)
2. Which hyper-parameter can help to reach the convergence point and the impact of value selection? (2 marks)
3. How can we find the global minima for the objective function shown in Figure-b? [N.B. Conceptual description will be accepted.] (3 marks)

Are you struggling to keep up with the demands of your academic journey? Don't worry, we've got your back! Exam Question Bank is your trusted partner in achieving academic excellence for all kind of technical and non-technical subjects.

Our comprehensive range of academic services is designed to cater to students at every level. Whether you're a high school student, a college undergraduate, or pursuing advanced studies, we have the expertise and resources to support you.

To connect with expert and ask your query click here Exam Question Bank