design and implement experiments.

NIT6130 – Introduction to Research Page 1 of 3 NIT6130 Introduction to Research Assignment 4 – Experiment Design & Result Analysis Semester 2, 2017 Due Date: 11:59pm, Friday, Week 12 Submission Requirement: submit online via a link in Assessment on VU Collaborate Value: 25% of semester assessment Objective: The purpose of this assignment is to help students getting familiar with the process of design and implement experiments for their research projects. Students will review, analyse, design, implement experiments to evaluate the proposed methodology in previous assignment, and report experimental results and findings as well. Tasks: 1. Collect your data for experiment a. Identify the available data sources for collection. b. Select the most appropriate data source(s) and start data collection. You are asked to record the data sources that you have found, brief description of the available data using the following sample table: Data Source Name Source Organization Data Description Data File Format URL (if available online) Charge Fee Target data source Data 1 Sport Centre, VU Exercise training record txt http://xxxxxxx Free Yes Data 2 … … … … $1000 No c. Store the collected data in appropriate file format. You need to save the collected raw data, and keep a copy at all times. i. Create a folder called “raw data” to store the collected raw data ii. Record your data collection using the following sample table: Data Source Name Date of Collection Saved File Location Saved File Name Saved File Format No. of Data Records Data 1 15/3/2016 //raw data/ xxxx.txt txt 2000 Data 3 … … … … NIT6130 – Introduction to Research Page 2 of 3 2. Design and implement experiments a. Adopt data pre-processing. Not all the collected raw data are available as direct input of your methodology. b. Feature selection or dimension reduction. You may need to select some of data features or data records from the entire data collection, not to use all the collected data. In other case, you may need to reduce the dimensionality of the collected data to simplify the data processing in the later stages. You need to save the result data set into a new file. You are asked to record these works using the following sample table. Date Data Source Name Purpose of Pre-processing Pre-processing Method No. Original Data Records No. Result Data Records No. Original Features No. Result Features New Data File Name 20/3/2016 Data 1 Clean the missing data Pre-fill the missing values 2000 2000 10 10 Data1_full.txt 22/3/2016 Data 1 Feature Selection xxxxxx 2000 2000 10 6 Data1_6features.txt … Data 3 … … … … c. Design your experiment based on the proposed methodology. The main purpose of experiments is to prove that your methodology can provide expected outcomes. You can follow the experiments that you have read from literatures or any other ways that is suitable for your case. Your experiments procedure need to be detailed recorded, including instruction steps, input data, expected output, potential problems and other related issues. When state the procedure of experiment, you can use extra table, figure, chart, diagram to provide better description. Date Experiment Purpose of Experiment Description of Procedure Input Data Expected Output Result File Name 26/3/2016 Experiment 1 Evaluate Method 1 … … … Output1.txt d. To implement your experiment, normally you need to use software/tools or write your own program. Record your experiment results for each run as the references of the following analysis task. 3. Result analysis and summary NIT6130 – Introduction to Research Page 3 of 3 a. What are the expected results that you want to obtained from the experiment? You need to have a rough idea of what you may have in advance; you cannot wait until the experiment is done. b. Plan your result analysis by thinking of what result you may have first. Write a summary of why you expected to have that results, how they can be linked to your research problems. 4. Write up your experiment and result analysis chapter. a. Provide an outline of your experiment and result analysis chapter, including up to four levels of subsections. Only need to list the chapter titles (level 1), section titles (level 2), subsection titles (level 3), and sub-subsection titles (level 4). Marking criteria Assessment Criteria: Mark Data Collection  Data source list  Data collection record 2 3 Experiment Design and Implementation  Data pre-processing  Feature selection or dimension reduction  Experiment design  Experiment implementation records 2 3 3 2 Experiment Result Analysis & Summary  The expected results  Summarize your expected results to answer research problems 2 3 Outline of Experiment and Result Analysis Chapter 5 Total 25

Leave a Reply

Your email address will not be published.