Machine Learning with Python
SIT720 Machine Learning Assessment Task 1: Problem solving task. ©Deakin University 1 SIT720 This document supplies detailed information on Assessment Task 1 for this unit. Key information • Due: Monday 26 July 2021 by 8.00 pm (AEST) • Weighting: 5% Learning Outcomes This assessment assesses the following Unit Learning Outcomes (ULO) and related Graduate Learning Outcomes (GLO): Unit Learning Outcome (ULO) Graduate Learning Outcome (GLO) ULO1 - Perform Python programming to solve a given problem. GLO1 - through the assessment of student ability to use data acquisition. techniques to obtain, manipulate and represent data. GLO2 - through the assessment of communicating the results in specific format. GLO3 - through student ability to use specific programming language and modules to obtain, pre- process, transform and analyse data. Purpose This assessment task is for student to apply Python programming skills for loading, visualising, manipulating and exporting data using various modules and packages. Assessment 1 Total marks = 30 15 * 2 = 30 Submission Instructions a) Submit your solution into a notebook file with “.ipynb” extension. b) Insert your Python code or text responses into the cell of your submitted file followed by the question i.e., copy the question by adding a cell before the solution cell. If you need multiple cells for better presentation of the code or answer, add question only before the first solution cell. c) Your submitted code should be executable. If your code does not generate the submitted solution, then you will get zero for that part of the marks. d) For answers regarding discussion or explanation, maximum five sentences are suggested. e) Answers must be relevant and precise. f) No hard coding is allowed. Avoid using specific value that can be calculated from the data provided. g) Submit your assignment after running each cell individually with the output. h) The submitted notebook file name should be of this form “SIT720_A1_studentID.ipynb”. For example, if your student ID is 1234, then the submitted file name should be SIT720_A1_1234.ipynb. Background According to World Health Organization (WHO), cardiovascular diseases (CVDs) are the number 1 cause of death globally, taking an estimated 17.9 million lives each year and affecting the quality of life of a large number of people worldwide. Prerequisites of the treatment of these types of disease involve proper diagnosis method to identify its occurrence and its type. SIT720 Machine Learning Assessment Task 1: Problem solving task. ©Deakin University 2 SIT720 Diagnosis of such diseases always involve a large number of parameters to help the Cardiologists to identify them. Electrocardiography (ECG) is most commonly used to observe patient's cardiac states. The following image shows different ECG waves (P, Q, R, S, T, etc). Fig: Electrocardiogram trace with respective biomarkers. [Image source: https://litfl.com/wp- content/uploads/2018/10/ECG-waves-segments-and- intervals-LITFL-ECG-library-3.jpg] In this assignment, you will have a look at such a dataset containing different parameters along with the decision of the Cardiologists about the level of a sample heart disease. There will be a list of tasks to check your ability to use of programming skill, basic logics, and reasoning. Dataset Dataset file name: A1_heart_disease_dataset.csv Dataset description: Dataset contains different features along with the disease state. It contains total 13 features and an additional disease state, in total 14 columns. It contains different types of data including int, float and string. Feature names, data type and values are described in the following section with their proper unit details. Data may contain 'null' or 'nan' values. Each observation is a datapoint along the row of the dataset. Patient and observation are used interchangeably in this case. Features: i. age (int): age of the patient in year ii. sex (str): gender of the patient (M: male, F: female) iii. cp (str): chest pain type (tap: typical angina, aap: atypical angina, nap: non-anginal pain, asp: asymptomatic pain) iv. trestbps (float): resting blood pressure (in mm Hg on admission to the hospital) v. chol (float): serum cholesterol in mg/dl vi. fbs (bool): is fasting blood sugar higher than standard 120 mg/dl? (yes: if true; no: if false) vii. restecg (int): resting electrocardiographic results (0: normal, 1: having ST-T wave abnormality (T wave inversions and/or ST elevation or depression of > 0.05 mV), 2: showing probable or definite left ventricular hypertrophy by Estes' criteria) viii. thalach (float): maximum heart rate achieved ix. exang (bool): exercise induced angina (true: if yes; false: if no) x. oldpeak (float): ST interval depression induced by exercise relative to rest %5bImage%20source:%20https:/litfl.com/wp-content/uploads/2018/10/ECG-waves-segments-and-%20intervals-LITFL-ECG-library-3.jpg%5d %5bImage%20source:%20https:/litfl.com/wp-content/uploads/2018/10/ECG-waves-segments-and-%20intervals-LITFL-ECG-library-3.jpg%5d SIT720 Machine Learning Assessment Task 1: Problem solving task. ©Deakin University 3 SIT720 xi. slope (float): the slope of the peak exercise ST segment (1: up-sloping, 2: flat, 3: down- sloping) xii. ca (int): number of major vessels (0-3) affected xiii. thal (int): thalassemia state (3: normal; 6: fixed defect; 7: reversable defect) xiv. state (int): heart disease risk state (0: no disease, 1-4: level of risk) [Decided by cardiologists] _____________________________________________________________________________________ Questions _____________________________________________________________________________________ 1. Load the data from supplied data file. Remove the observations/samples where the heart diseases are not diagnosed by the Cardiologists. Print the data dimension before and after removing the observations/samples. 2. Continue from question 1. Display the number of rows and their indices that have missing data in one or more cells. Now, replace the missing data by the lowest value of the corresponding feature if it is a continuous variable. In case of categorical variable, remove the sample. Print the median values of all features before and after replacing missing data. 3. Continue from question 2. Is there any change in data type? If yes, convert them back to appropriate data types. Print all variables with corresponding data type. 4. Continue from question 3. Print the total numbers and ration of male and female patients who are at highest risk of heart disease. 5. Continue from question 3. Is there any association between heart rate and severity of heart disease? Explain your results from given dataset. 6. Continue from question 3. Print the average cholesterol level for different number of blocked blood vessels across gender. Please report the pattern found in the result, if any. 7. Print the percentage of patients at risk of heart disease having abnormality in both ECG and blood sugar with asymptomatic chest pain. 8. Calculate and print the average blood pressure of all observations with non-flat ST slopes of ECG. 9. Create and print a dataframe of the heart rate, blood pressure and cholesterol levels for different age groups (based on 10 years interval). 10. Continue from question 3. Find the average cholesterol level of across gender for each age group. Please explain the results. 11. Continue from question 3. Draw two scatter plots of cholesterol level, one against blood pressure and another against heart rate. Draw them in two subplots of the same plot. 12. Visualize the cholesterol level against number of blood vessel blocked for male and female using line plot. Explain the graph base on your observation. 13. Draw a group bar diagram of heart rate, blood pressure and total number of patients, based on age groups defined in question 9. Explain your observation from the graph. SIT720 Machine Learning Assessment Task 1: Problem solving task