Study Guide

Objectives (course unit)

The student understands basic concepts of AI and Machine Learning. The student is able to create and use Machine Learning Algorithms in Python. The student learns how to make analysis and predictions and knows which Machine Learning model to choose for each type of a problem.

Content (course unit)

- Basic concepts of AI and Machine Learning
- Unsupervised and Supervised learning
- Regression, Association, Classification
- Naïve Bayes, Decision Trees and Neural Network Algorithms
- Training and validation of models
- Production testing of models

Prerequisites (course unit)

Basic knowledge of programming

Further information (course unit)

Includes content of previous Mathematics 3 course. The course eliminates duplication observed in courses.

Assessment criteria, satisfactory (1-2) (course unit)

Student knows about the basic concepts of AI and Machine Learning. Student can apply at least some supervised or supervised learning applications. Student can use regression, association or classification algorithm with support. Student can create an application using either Naïve Bayes, Decision Trees or Neural Network Algorithms. Student can setup training and validation processes for new models. Student can setup production testing for new models.

Assessment criteria, good (3-4) (course unit)

Student knows and understands the basic concepts of AI and Machine Learning. Student can apply both supervised and supervised learning applications. Student can create applications with regression, association, or classification algorithms. Student can create working applications using Naïve Bayes, Decision Trees and Neural Network Algorithms. Student can setup and apply training and use validation methods for new models. Student can follow procedures of production testing for new models.

Assessment criteria, excellent (5) (course unit)

Student knows and understands in depth the basic concepts of AI and Machine Learning. Student can apply both supervised and supervised learning for various applications. Student can use regression, association, and classification algorithms where appropriate. Student can create versatile applications using Naïve Bayes, Decision Trees and Neural Network Algorithms. Student can implement various training and validation solutions for new models. Student is able to execute reliable production testing for new models.

Location and time

AI & ML: 3 hours per week in classroom
Mathematics: 1 hour online, 2 hours in classroom per week (5 weeks total)

Exam schedules

No exam.

Retake and improvement of the grade :
First retake on week XX/2024. Second retake on week YY/2024. A student contacts the lecturer during the retake week for detailed instructions. Improvement of the grade can be tried once during the retake weeks.

Math part exams and retakes (TENTATIVE) on weeks 44, 48 and 50. To be confirmed by the first math part classes.

Assessment methods and criteria

The course consists of two separate parts: ML&AI and Mathematics. A student gets a separate grade from both parts. The final course grade is weighted average of the grades of the parts. ML&AI is 5/8 of the final course grade and Mathematics is 3/8 of the final course grade.

ML&AI:
A student can get points from two separate final practical works. Max. points for Practical work 1 is 20 points. Max. points for Practical work 2 is 30 points.

ML&AI points and grades:
0 0
10 1
17 2
25 3
35 4
45 5

--------------

Mathematics:
The scores in Mathematics part are received from continuous proof (50%) and a supervised exam (50%)

Mathematics part grade thresholds are:
30% 1
45% 2
60% 3
75% 4
90% 5

Student must receive at least 20% of maximum score in a supervised exam to have approved grade of the math part

If student utilizes ready-made artificial intelligence tools (e.g. MS Co-pilot, ChatGPT) in course assignments, the student must give reference to which AI tools have been used and report the prompts the student has used. Student must be able to narrate the exercise solutions submitted by the student. Teacher has a right to ask student, whether artificial intelligence tools have been used and require student to complete their assignment, if AI tools have been used inadequately.

Assessment scale

0-5

Teaching methods

AI & ML: 3 hours per week in classroom
Mathematics: 1 hour online, 2 hours in classroom per week (5 weeks total)

Learning materials

Course materials in Moodle:
https://moodle.tuni.fi/course/view.php?id=44865

Student workload

75 hours contact teaching and 138 hours independent learning.

Content scheduling

Course schedule is in course Moodle.

Course content:

Basics of Machine Learning and AI
Linear Regression
Logistic Regression
Decision Tree
Random Forest
ANN
CNN
Mathematics

Remark: Selected mathematical concepts will be introduced in math part of the course before they are applied in the analysis part

Completion alternatives

To be negotiated with the teacher responsible for the course part. Teacher is not obliged to grant alternative ways of completion