Assessment The course assessment is divided into 80% assignment, 10% presentation, and 10% seminar paper. 80% Assignment You can work in groups of two. Unless otherwise specified, equal contribution will be assumed The 80 points are divided into 80% for the quality of the project and 20% for the quality of the demonstration The aim of this project is to introduce you to machine learning techniques applied to real data sets. You are expected to: Choose a data set from the repository in http://archive.ics.uci.edu/ml/ Compare at least 2 machine learning techniques for the modelling and prediction of data in the chosen data set In your experiments you must make use of the following techniques rescaling and normalization cross validation dimensionality reduction feature selection evaluation method Write and hand in a report where you explain all the steps of the implemented methodology and experiments carried out. Skeleton of the report Section 1 - Introduction Explain the properties of the chosen data set and what you will be doing with it Mention the two (or more) machine learning techniques that you will be using Section 2 - Background Describe the mechanics of the selected machine learning techniques Describe what rescaling and normalisation are and why they are important Describe what cross validation is Describe what dimensionality reduction and feature selection methods are Explain the quantitative measurements that you will be using to quantify the results; e.g. accuracy rate Section 3 - Experiments Describe the steps that you used to process the data set Describe the experiments that you carried out Section 4 - Conclusions Draw conclusions from your experiments. Example feature selection with entropy works better than PCA, ... Citations When you use citations, please use the IEEE standard referencing style. Formatting of Report It should not be longer than 20 A4 pages including everything Single line spacing Font type Arial, font size 10 Deadline Report must be handed in to Ms. Francelle Scicluna (Level 1, Block A, Room 4) not later than 12:00h Wednesday January 13. It is a hard deadline. There will be 10% deducted with every one hour of delay. 10% Presentation Every student is expected to present a paper about a machine learning technique. The paper must first be sent by email to Dr. George Azzopardi to confirm its appropriateness, by not later than 17:00h Monday January 4. The presentation will take 15 minutes + 5 minutes discussion All presentations will be given during the session of Wednesday January 6. An alphabetical order will be considered. The 10 points are divided as follows: 50% - average of the grades anonymously collected from the audience (i.e. your fellow students) 40% - quality of your presentation assessed by the lecturer 10% - active participation during the discussions of other studentsÂ’ presentations 10% Seminar Paper Choose a paper from the IDA proceedings of 2015 or from any other suitable conference proceedings or journals. Write a review about the selected paper. The review must consist of a summary of the paper and constructive criticism of the proposed work. Not longer than one A4 page. Font type: Arial, font size: 10.