SCHOOL OF COMPUTER SCIENCE
MASTER OF APPLIED COMPUTING (MAC)
ASSIGNMENT 3 (Weightage 30%)
SEPTEMBER 2024 SEMESTER
|
MODULE NAME
|
: Principles of AI
|
|
MODULE CODE
|
: ITS70304
|
|
DATE/TIME
|
: 8:00 PM (MYT)
|
|
PLATFORM
|
: TIMeS
|
Assessment Criteria
|
Assessment
Task
|
Weightage
|
MLO Assessed
|
Formative/
Summative
|
Assessment Instrument
|
Topics
|
Week
|
MQF 2.0
|
|
Part I
|
15%
|
3
|
Formative
|
N/A
|
3,4,5,6
|
3,4,5,6
|
C1
|
|
Part II
|
15%
|
4
|
Google Collaboratory
|
C3A
|
MLO3 - Proposed and select suitable AI or Machine learning algorithm for a given application.
MLO4 - Analyze an AI-based solution for a given application.
C1 = Knowledge & Understanding, C2 = Cognitive Skills, C3A = Practical Skills, C3B = Interpersonal Skills, C3C = Communication Skills, C3D = Digital Skills, C3E = Numeracy Skills, C3F = Leadership, Autonomy & Responsibility, C4A = Personal Skills, C4B = Entrepreneurial Skills, C5 = Ethics & Professionalism
Scenario and Task Description
Predictive analytics in healthcare uses machine learning techniques and statistical algorithms to process real-time and historical patient data and discern patterns, trends, and associations for accurate predictions of individual health outcomes and population-wide events.
Itransition’s experts use their robust experience in predictive analytics software development and implementation to create custom solutions tailored specifically for healthcare providers in line with the latest industry trends and regulations.
The numerous benefits of predictive analytics for healthcare organizations, individual patients, and the population make it one of the most used types of analytics in the industry.
Predictive analytics helps healthcare professionals spot trends in patients’ medical records and genomic information that are impossible to detect manually, helping to understand patient conditions better. Predictive modeling, in particular, allows forecasting the course of the disease, enabling medical professionals to avoid health risks like adverse reactions to medicine, genetically determined resistance to treatment, and failure to adhere to the regimen. This way, clinicians receive actionable insights that help them optimize patients’ treatment plans according to their physical and psychological needs.
Natural Language Processing (NLP) is a subfield of artificial intelligence (AI) that focuses on the interaction between computers and humans through natural language. The goal of NLP is to enable machines to understand, interpret, and generate human language in a way that is both meaningful and useful. It is a multidisciplinary field, drawing from computer science, linguistics, and cognitive psychology. NLP can be used to as a predictive modelling in healthcare.
Practical Skills: Part I
1. Describe FOUR different techniques of NLP (Rule based, Statistical Methods, Machine Learning and Deep Learning). Ensure to relate your description with the usage of NLP. (2 marks)
2. Data cleaning step is crucial to ensure that you’re working with high-quality data. A good data scientist will spend around 70-90% of their time cleaning their data. Use Pandas and NumPy libraries to perform. some of the required data cleaning steps. (you can include data normalization-encoding and scaling if necessary) (5 marks)
3. Carrying out a simple Exploratory Data Analysis (EDA) from the healthcare.csv dataset. Predicted column has three values (Normal," "Abnormal," or "Inconclusive”). Perform. a way to make the predicted column into only a binary classification. (3 marks)
4. When your data is ready for modelling, you can start building your prediction model. Perform. TWO (2) modelling (please use only the modelling covered in the module). Justify the selected Machine Learning models and describe them. (5 marks)
Practical Skills: Part II
5. There are TWO (2) basic matrices to measure a performance from a Machine Learning model. Describe those TWO (2) matrices. (4 marks)
6. Evaluate the performance for both Machine Learning models you used in Part I. Show their accuracy, precision and recall values. Explain the results you have received for both models.
(8 marks)
7. Prepare a sample of test data and show how the best model you built can predict whether the test data can give positive or negative emotion. (3 marks)
To demonstrate a broad and coherent theoretical and technical knowledge comprehension,
add comments where necessary throughout the program. Please make sure you copy paste the respective code in your pdf file and explain each of them.
|
Marking Rubrics (lecturer’s use only)
Attach as second page in the report.
|
|
The purpose of this learning assignment is based on the following module learning outcome (MLO):
MLO3 - Proposed and select suitable AI or Machine learning algorithm for a given application.
MLO4 - Analyze an AI-based solution for a given application.
Type of activity: Practical
|
|
Question
|
Weight
|
Outstanding
(80 – 100)
|
Mastering
(65 – 79)
|
Developing
(0 – 64)
|
|
Part I
|
Accurately describe Natural Language Processing and show clear understanding and generation in it. The similarity is less than 2%.
|
Accurately describe Natural Language Processing but not clearly show the understanding and generation in it. The similarity is between 2% to 4%.
|
Not accurately describe Natural Language Processing and no understanding and generation in it. The similarity is greater than or equal to 5%.
|
|
Q1
|
_____/2
|
|
Q2
|
_____/5
|
Correctly perform. Data cleaning and Data Normalization. The similarity is less than 2%.
|
Correctly perform. Data cleaning and Data Normalization but no comprehensive explanation on both processes. The similarity is between 2% to 4%.
|
Incorrectly perform. Data cleaning and Data Normalization but no comprehensive explanation on both processes The similarity is greater than or equal to 5%.
|
|
Q3
|
_____/3
|
Correctly perform. Exploratory Data Analysis (EDA) and perform. the binary classification conversion. The similarity is less than 2%.
|
Adequate perform. Exploratory Data Analysis (EDA) and perform. the binary classification conversion. The similarity is between 2% to 4%.
|
Incorrectly perform. Exploratory Data Analysis (EDA) and wrongly perform. the binary classification conversion. The similarity is greater than or equal to 5%.
|
|
Q4
|
_____/5
|
Demonstrates comprehensive steps to build the two modelling with correct justification. The similarity is less than 2%.
|
Demonstrates comprehensive steps to build the two modelling adequately with adequate justification. The similarity is between 2% to 4%.
|
Did not demonstrate comprehensive steps to build the two modelling and incorrectly justify the steps. The similarity is greater than or equal to 5%.
|
|
Part II
|
Describe those TWO (2) matrices correctly. The similarity is less than 2%.
|
Describe those TWO (2) matrices adequately. The similarity is between 2% to 4%.
|
Describe those TWO (2) matrices incorrectly. The similarity is greater than or equal to 5%.
|
|
Q5
|
_____/4
|
|
Q6
|
_____/8
|
Their accuracy, precision and recall values are calculated correctly and the comparison between those two models’ performance are explained precisely. The similarity is less than 2%.
|
Their accuracy, precision and recall values are calculated correctly but the comparison between those two models’ performance is explained adequately. The similarity is between 2% to 4%.
|
Their accuracy, precision and recall values are calculated incorrectly and the comparison between those two models’ performance are explained wrongly. The similarity is greater than or equal to 5%.
|
|
Q7
|
_____/3
|
Correctly demonstrate sample of test data and show how the best model built can predict whether the test data can give positive or negative emotion. The similarity is less than 2%.
|
Adequately demonstrate sample of test data and show adequately how the best model built can predict whether the test data can give positive or negative emotion. The similarity is between 2% to 4%.
|
Incorrectly demonstrate sample of test data and can’t show how the best model built can predict whether the test data can give positive or negative emotion. The similarity is greater than or equal to 5%.
|
Submission Requirements
1. Font type : Times New Roman
2. Font size : 12
3. Line spacing : 1.5
4. Alignment : Justify Text
5. Document type : .pdf, .ipynb
6. Number of pages : 5 – 12 pages (do not exceed the page limit)
7. Your full report should consist of the following:
a) Cover page (Name, ID, Date, Signature, Score)
b) Marking Rubrics & Declaration (attach as second page in the report)
c) Report of your answer script.
d) Appendixes (line spacing = 1.0)
· List of references (APA format)
· Python script.
· Report of similarity score (percentage of similarity score from each source needs to be shown)
8. Start each question on a separate page.
9. All figures and tables are labelled properly.
10. File naming conventions: StudentName_Assignment1
Notes:
· Student is not allowed to transcribe directly (copy and paste) any material from another source into their submission.
· Start each question on a new page.
· Answer in form. of short essay (50 to 200 words) and print out the relevant Python program outputs
· All process/functions must be clearly explained.
· Include in-text citation to support your answers and add the list of references at the end of your report (APA format). The list of references is to be alphabetized by the first author's last name, or (if no author is listed) the organization or title.
· The Turnitin similarity for this module is 20% overall and lesser than 1% from a single source excluding program source codes.