ML Project - M1 - ISII - 2021 2022 1 1 2 1 1

Machine Learning M1/ISII 1
University of Algiers - 1 Benyoucef Benkhedda Academic Year 2022/2023

Faculty of Sciences Module: Machine Learning
Department of Computer Science
ML Mini
Project
Project Objective:
The purpose of this project is, in the first place, to familiarize you with the structure, selection, and quality of
datasets for supervised learning, including Regression and Classification issues. In the second place, to enable
you to deepen your understanding of machine learning, by studying and applying different supervised learning
techniques, seen in class, on the selected regression and classification problems.
The work MUST be done in teams of five students, which means 5 persons make a 'Team' to work together on
this project. The team’s members can be from different TP groups.
 You are asked the following:
a) Go to the Machine Learning Repository of the University of California at Irvine (UCI).

https://archive.ics.uci.edu/ml/datasets.php
b) Start with a quick site scan to understand what it is. On the left side of the page, you can see different tasks
including regression and classification.
c) Select from this website:
- Three (3) training data for the regression, and

- Three (3) training data for the classification.
 Whatever problems you select, make sure that they are appropriate for regression and classification techniques,
and that they are not very complex or very simple problems. It's up to you to choose the datasets that suit you
the most.
d) Introduce on the “Machine Learning” course platform (in order of priority of problems):
- The full names of your team members,

- The 3 data choices for the regression task,
- The 3 data choices for the classification task.
e) The submission of team names and problem choices on Google drive will be closed on
8/12//2022 at 11. 59p.m (23:59 - Be punctual!)

Start the Project :
1) Do the necessary research on your two assigned problems and apply the necessary data preprocessing to
improve its quality, its representation if it is modified to be more appropriate to the techniques considered,
etc.
2) Design your solution for each supervised learning technique covered in class and implement it to solve these
problems using Python (Python 3 is recommended). Apply the techniques implemented on your datasets that
will be used to train your model:
 Linear Regression (LnR)
 Logistic Regression (LgR)
 Neural Networks (NN)
 Support Vector Machine (SVM)
 K-nearest neighborhood (knn)
 Gradient Descent (GD)
 Normal Equation (NE)
 Regularization, Cross Validation, …
3) Prepare a report of no more than 20 pages (without the appendix), Times New Roman, 11, single spaced
(paragraph format) in English or French languages. This report will contain two main parts, a part for
the regression task and a part for the classification task. For each part you must:
-Present the problem in question in detail with its characteristics.

-Present the selected data set, explaining the reasons for this choice. Present the characteristics of the database
such as its size, the number and types of attributes with their meanings, the different problems present in the
database with examples
-Present a clear explanations of the methods applied to improve its quality
-Any other information that you find useful and necessary related to your problems being treated.
-Explains with all the necessary details and justify your design choices of the best architecture and
configuration for each technique and the results obtained, including summary / comparison tables, graphs,
analyzes, discussions of results, conclusions, etc.
4) Prepare a video presentation of the entire Mini Project, including the demo of the solutions. Each team will
present the proposed solutions in video (record).. You must have a good grasp of the problems, the techniques
and the various details of the solution design.
 What you must submit for each team:
a. Each team must submit a report respecting the number of pages and the format indicated above, and
clearly explaining what was requested.
b. The report must be sufficiently complete .You are limited to 20 pages in Times.
c. An online submission through Google drive
3 files: a- The Report (Two electronic versions (pdf +

word)) +record or video presentation.
b- The Presentation (video or record)
c- A rar / zip file that contains:
 The report of your mini project pdf and word versions

 The processed datasets used for both Regression and Classification problems +links of datasets
 All the Scripts of your solutions well organized in directories and having meaningful names. Use
meaningful names for variables and functions in your implementation as well as clear and
precise comments in the code to clarify your work.
 All screenshots of figures from your tests with meaningful names. The report should contain
the comparison tables of the results and the figures that you think are important.
d. « Non-plagiarism and intellectual dishonesty commitment » downloaded from the platform, completed,
signed by each member of the team, and inserted in your report at the end. Certifying that your reports as
well as your code for the whole project is entirely your work except where references are given (without
copy / paste).
Learn to rely on your personal efforts and develop your skills and intellectual honesty!
 Important dates :
- Completed Work Submission Deadline:09/01/2023 11.59 pm
NB
1. Work that is not submitted through the Google drive will not be accepted.
2. The names of Scripts, Figures and Reports must contain the team number « E.g.: X_Team30.(ext) »
3. Only one submission per team, which means, one person will be responsible for submitting the names
and team selections for the first phase, as well as the electronic versions for the second phase.
4. Each double submission results in loss of points for the team.
5. Each late submission results in loss of points for the team.
6. Each plagiarism detection will result in loss of points, or even cancellation of the project, for the
team, except where references are given without copying / pasting (an anti-plagiarism tool will be used
to verify that it is of your work). No form of phrase and / or code plagiarism (without references)
will be tolerated.
7. The grade for your work will take into consideration the quality of the work; compliance with the requested
formats; the respect of deadlines and organization of your project.!
8. Any interesting innovation will be rewarded.
9. The score of each member of the same team will not necessarily be the same, each according to his/her
contribution and understanding of the problem, techniques, solutions,, answers to questions during the
presentation, etc.…
Good luck

ML Project - M1 - ISII - 2021 2022 1 1 2 1 1

Uploaded by

Copyright:

Available Formats

ML Project - M1 - ISII - 2021 2022 1 1 2 1 1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ML Project - M1 - ISII - 2021 2022 1 1 2 1 1

Uploaded by

Copyright:

Available Formats

Machine Learning M1/ISII 1

University of Algiers - 1 Benyoucef Benkhedda Academic Year 2022/2023

 You are asked the following:

a) Go to the Machine Learning Repository of the University of California at Irvine (UCI).

c) Select from this website:

- Three (3) training data for the regression, and

- The full names of your team members,

8/12//2022 at 11. 59p.m (23:59 - Be punctual!)

-Present the problem in question in detail with its characteristics.

 What you must submit for each team:

c. An online submission through Google drive

3 files: a- The Report (Two electronic versions (pdf +

 The report of your mini project pdf and word versions

- Completed Work Submission Deadline:09/01/2023 11.59 pm

4. Each double submission results in loss of points for the team.

5. Each late submission results in loss of points for the team.

8. Any interesting innovation will be rewarded.

You might also like