Page 1 KIT108 Artificial Intelligence Generated by Akari KIT108 ARTIFICIAL INTELLIGENCE Assignment Stage 1 submission: Due 23.59PM Wednesday 27th May 2020 (already includes a 5 day extension) Stage 2 submission: Due 23.59PM Sunday 07th June 2020 (already includes a 5 day extension) STOCK PRICE PREDICTION Description In this assignment, we will apply machine learning techniques learned in the lectures and tutorials to predict the highest price of a stock in the next day. 1. The date – “Date” 2. The opening price of the stock – “Open” 3. The low price of that day – “Low” 4. The closed price of that day – “Close” 5. The amount of stocks traded during that day – “Volume” Page 2 KIT108 Artificial Intelligence Generated by Akari 6. The high price of next day – “Next High” Part A: Programming- 70% In this task, you will train an AI model using the open price, the low price, and the volume of a day to predict the high price on the next day. In this task you need to design a RapidMiner process OR Python program to: 1. Read the stock_data.csv file. 2. Identify irrelevant information from the data and filter it out to construct the target data. Explain in the report how you do this. 3. Identify the number of missing values in each attribute. Explain in the report. 4. Fill the missing values by using the techniques you have learned. Explain the way you handled this issue in the report. 5. Normalise the data so that all attributes are in the same range. Explain what method you used and what your chosen range was in the report. NOTE: DON’T NORMALISE THE TARGET (LABEL) ATTRIBUTE. 6. Decide your own strategy to train, evaluate a model from the data from stock_data.csv. 7. Design your own strategy to select the best model. 8. Apply the selected model to the data in predict_stock_data.csv. This data does not have the label and you have to generate the predicted value for the high price. Export the predicted high price in to a csv file using your student ID (for example 342435.csv). When you complete the step 8, you will do stage-1 submission in week 13 (see below) to score the performance of your model. Part B: Analysis- 30% You will receive the performance score after week 13 and the ground truths. You will need to revise your design to: – Explain why your model is good or bad. Write it in the report. – What will you do to improve the prediction results if your model is scored as “NOT GOOD”. Write this in the report. Those whose models are scored as “GOOD” do not need to do this step. Page 3 KIT108 Artificial Intelligence Generated by Akari Evaluation of the assignment See Page 7 and Page 8 for the details of the evaluation criteria. How & What to submit. Assignments will be submitted via MyLO (an Assignment submission will be created). Stage 1 submission (week 13): • An XML file for your design OR a python source code, the file is named using your student ID (for example, 342435.xml or 342435.py) • The predicted output in a csv format. The file must be named using your student ID (for example 342435.csv). After stage 1 submission, you will receive the score (GOOD or NOT GOOD) for the model, from which you will complete a report for the stage 2 submission. Stage 2 submission (week 15): • A report file (docx or pdf) using the provided template (see the attached file) Appendix 01: How to export an XML file for the model with RapidMiner • Click File -> Export Process • Choose save path -> Name file with ID -> Select Process File (*.xml) NOTE: YOU NEED TO CHECK YOUR XML FILE BEFORE SUMISSION, FOR EXAMPLE BY CLOSING RAPIDMINER STUDIO THEN OPENNING IT AGAIN AND LOADING THE XML FILE FOR DOUBLE CHECK. Page 4 KIT108 Artificial Intelligence Generated by Akari Appendix 02: How to export an csv file for the predicted output using RapidMiner • Use Select Attributes operator to filter output with only predicted high price attributes left • Put Write CSV operator and connect with the filtered output • Set csv file path • Use symbol “,” for column separator • Unselect quote nominal values • Run process and output csv file Plagiarism and Academic misconduct Plagiarism Plagiarism is a form of cheating. It is taking and using someone else’s thoughts, writings or inventions and representing them as your own; for example, using an author’s words without putting them in quotation marks and citing the source, using an author’s ideas without proper acknowledgement and citation, copying another student’s work. If you have any doubts about how to refer to the work of others in your assignments, please consult your lecturer or tutor for relevant referencing guidelines. You may also find the Academic Honesty site on MyLO of assistance. Page 5 KIT108 Artificial Intelligence Generated by Akari The intentional copying of someone else’s work as one’s own is a serious offence punishable by penalties that may range from a fine or deduction/cancellation of marks and, in the most serious of cases, to exclusion from a unit, a course or the University. The University and any persons authorised by the University may submit your assessable works to a plagiarism checking service, to obtain a report on possible instances of plagiarism. Assessable works may also be included in a reference database. It is a condition of this arrangement that the original author’s permission is required before a work within the database can be viewed. For further information on this statement and general referencing guidelines, see the Plagiarism and Academic Integrity page on the University web site or the Academic Honesty site on MyLO. Academic misconduct includes cheating, plagiarism, allowing another student to copy work for an assignment or an examination, and any other conduct by which a student: a. seeks to gain, for themselves or for any other person, any academic advantage or advancement to which they or that other person are not entitled; or b. improperly disadvantages any other student. Students engaging in any form of academic misconduct may be dealt with under the Ordinance of Student Discipline, and this can include the imposition of penalties that range from a deduction/cancellation of marks to exclusion from a unit or the University. Details of penalties that can be imposed are available in Ordinance 9: Student Discipline – Part 3 Academic Misconduct. Page 6 KIT108 Artificial Intelligence Generated by Akari KIT108 ARTIFICIAL INTELLIGENCE: MAJOR ASSIGNMENT Synopsis of the task and its context This is an individual assignment making up 20% of the overall unit assessment. The assessment criteria for this task are: 1) Apply machine learning pipeline to solve a real-world problem. a) Identify relevant data b) Process and clean data c) Transform data d) Apply and select machine learning techniques e) Analysis of the results. f) Identify the best technique for this problem. Match between learning outcomes and criteria for the task: Unit learning outcomes On successful completion of this unit… Task criteria: 1. understand the local and global impact of AI on individuals, organizations, and society 2 2. adapt and apply techniques for acquiring, representing, and reasoning with data, information, and knowledge 1 3. select and effectively apply techniques to develop simple AI solutions 1 4. analyze a problem, apply knowledge of AI principles, and use ICT technical skills to develop potential solutions 1, 2 5. evaluate strengths and weaknesses of potential AI solutions 1, 2 Page 7 KIT108 Artificial Intelligence Generated by Akari Criteria HD (High Distinction) DN (Distinction) CR (Credit) PP (Pass) NN (Fail) 1. Machine learning pipeline a) Data collection (10%) Load the data and choose the irrelevant attribute (date) and be able to perform removal of that attribute from the original dataset. Load the data and remove the irrelevant attribute date, but also identify and remove one attribute
wrongly. Load the data and remove the irrelevant attribute date, but also identify and remove more than one attribute wrongly. Load but cannot remove the irrelevant attribute. Cannot load the data. b) Data processing (10%) Fill all the missing values and explain clearly the technique used to do that. Fill all the missing values and can explain technique used to do that with minor mistakes. Fill all the missing values but cannot explain the techniques correctly. Fill all the missing values but do not explain the techniques (no attempt). Cannot fill all the missing values. c) Data transformation (10%) Correctly normalise all the attributes into a range, except for the target attribute (next high). Correctly normalise all the attributes into a range but mistakenly normalise the target attribute (next high). Correctly normalise some of the attributes into a range, except the target attribute (next high). Correctly normalise some of the attributes into a range but mistakenly normalise the target attribute (next high). Cannot normalise the data. d) Data Mining (30%) Correctly use more than 2 models to select the best model to generate a prediction result from the predict_stock_data.scv. The selection process is explained correctly. Correctly use more than 2 models to select the best model to generate a prediction result from the predict_stock_data.scv. The selection process is explained with some mistakes. Correctly use more than 1 models to select the best model to generate a prediction result from the predict_stock_data.scv. The selection process is explained with some mistakes. Correctly use only 1 model to generate a prediction result from the predict_stock_data.scv. Cannot apply a model for the prediction. e) Pattern Evaluation (10%) Correctly design the evaluation step and choose a relevant evaluation metric. Design the evaluation step and choose a relevant evaluation metric with some minor mistakes. Design the evaluation step but using a wrong evaluation metric. Design a wrong evaluation step and use a wrong evaluation metric but the system can still run. Cannot evaluate a model. 2. Analysis (30%) Page 8 KIT108 Artificial Intelligence Generated by Akari a) Explain why the submitted model works well/badly (15%) Explain correctly why the selected model gave good or bad results using the evaluation score and the ground truth given after the stage 1 submission. Explain with minor mistakes why the selected model gave good or bad results using the evaluation score and the ground truth given after the stage 1 submission. Explain why the selected model gave good or bad results, partly using the evaluation score and the ground truth given after the stage 1 submission. Explain why the selected model works gave good or bad results BUT NOT using the evaluation score and the ground truth given after the stage 1 submission. Cannot provide any explanation. b) Make a GOOD model (15%) Already have the GOOD evaluation score after stage 1 submission. OR Explain correctly how to improve the performance using the ground truth. Provide evidence for the explanation (i.e. new results) Explain correctly how to improve the performance using the ground truth. Evidence for the explanation is not provided. Explain how to improve the performance using the ground truth with minor mistake. Attempt to explain how to improve the performance. No attempt
