International Teaching | IOT DATA ANALYTICS
International Teaching IOT DATA ANALYTICS
Back
cod. 0522500131
IOT DATA ANALYTICS
| 0522500131 | |
| COMPUTER SCIENCE | |
| EQF7 | |
| COMPUTER SCIENCE | |
| 2024/2025 |
| YEAR OF DIDACTIC SYSTEM 2016 | |
| SPRING SEMESTER |
| SSD | CFU | HOURS | ACTIVITY | |
|---|---|---|---|---|
| INF/01 | 9 | 45 | LESSONS |
| Objectives | |
|---|---|
| The goal of this course is to provide students with methodological and technological skills to analyze in real time big data streams generated by IoT devices. Thus, the course aims to complement skills acquired during a bachelor level database course with skills pertaining the extraction of sensor data, the assessment of their quality, the analytical models and the machine learning techniques for data streams. Knowledge and understanding: Provide the student with knowledge on the models and the technologies to manage big data streams generated by IoT devices, aiming to trigger analytical processes useful for the optimal management of productive processes and technological networks. More specifically, the course aims to provide students with the following skills: -Sensor data processing -Specific analytics models used in IoT -Machine learning for IoT -Sequence Data Mining -Analysis of Time Series Applying knowledge and understanding: The course aims to provide students with the following abilities: • Know how to extract, manage, and process big data streams generated by IoT devices • Know how to analyze and improve in real time the quality of data generated by IoT devices • Know how to select specific analytical and machine learning techniques suitable for analyzing sensor and IoT data. |
| Prerequisites | |
|---|---|
| Students should be familiar with fundamentals of data management, distributed systems, object-oriented paradigm, and a programming language. |
| Contents | |
|---|---|
| After introducing IOT systems and the new application scenarios related to the management of big data streams generated from IOT devices, the course will focus on the following topics: Introduction to IOT Data Analytics (4 hours of theory) • Big data issues in the IOT context (2 hours of theory) • Case Studies (2 hours of theory) Real-time sequence mining (10 hours of theory) • The Stream Data model (1 hour of theory) • Sampling data streams (1 hour of theory) • Filtering streams: The Bloom filter (1 hour of theory) • Counting Distinct Elements in a Stream (1 hour of theory) • Estimating Moments (1 hour of theory) • Counting elements in a Window of a Stream (2 hours of theory) • Decaying windows (2 hours of theory) • Mining sequencial patterns (1 hour of theory) Machine learning for IoT (13 hours of theory) • Taxonomy of Machine Learning Systems (3 hours of theory) • Construction of Machine Learning Systems (6 hours of theory) • Clustering data streams (4 hours of theory) Data Serie Analytics (10 hours of theory) • Introduction to Data Series (1 hour of theory) • Data Series representation (2 hours of theory) • Distance Measures for Data Series (2 hours of theory) • Lower Bounding (1 hours of theory) • Indexing Data Series (2 hours of theory) • Analysis Methods (2 hours of theory) Tools for IoT Data Analytics (8 hours of lectures) • The Python language (4 hours of lectures) • Online Learning Frameworks and Tools (4 hours of lectures) |
| Teaching Methods | |
|---|---|
| The course includes 37 hours of lectures on theoretical topics and 8 hours on tools, aiming to introduce concepts and to develop abilities to design and implement solutions for real-time analysis of data streams originated from IOT systems and devices. Course contents are presented through powerpoint slides, stimulating critical discussions with the students. For each presented topic, the instructors will illustrate potential tasks on which a student or a group can develop the course project. As for tools, other than powerpoint slides, through which concepts and possible additional resources, such as links to forums, manuals, and other sites are presented, students are given the possibility to ask support, during office hours, on simulations they performed on their personal computer, to ask clarifications, and solve possible technical problems with the assistance of the instructors. |
| Verification of learning | |
|---|---|
| The achievement of the course objectives is certified by means of an exam, whose final grade is expressed on a scale of 30. The exam consists of a written test (student might be exempted by passing a midterm written test, if organized), and an oral examination. Optionally, students might develop a course project. The written test (or the midterm test) aims to assess the acquisition of the theoretical concepts presented during the course. The oral test must always be the last one. It consists of an interview, with questions on the theoretical and methodological contents taught during the course, aiming to assess the level of knowledge and understanding, as well as the ability to expose concepts. Through the optional project, students might show their ability to apply the acquired knowledge in real scenarios, and can carry it out individually or in groups of up to 3 students, who can choose from a range of proposals provided by the instructors. During the project development, students can interact with the instructors in order to communicate the project’s progress and possible critical issues, debating on the goals of the project and the modalities to continue it. At the end of the project, students should deliver a technical report containing the project documentation, and make a powerpoint presentation lasting about 30 minutes, which can be done together with the oral examination or before this. The final grade is assigned through the average of the grades on a scale of thirtieths reported on the written and the oral examinations. This grade can be increased by 1 to 3 marks by with the optional course project. |
| Texts | |
|---|---|
| 1.Jure Leskovec, Anand Rajaraman, Jerey D. Ullman, “Mining of Massive Datasets”, 3^ Edizione, Cambridge University Press, 2020. 2.Aurélien Géron, " Hands-On Machine Learning with Scikit-Learn, Keras, and Tensorflow: Concepts, Tools, and Techniques to Build Intelligent Systems “, 2^ Edizione, O Reilly ed, 2019. 3.J. Gama, Knowledge Discovery from Data Streams, CRC PRESS, 2010. 4.A. Bifet, R. Gavaldà, G. Holmes, B, Pfahringer, Machine Learning for Data Streams – with practical examples in MOA, The MIT Press, 2018. |
| More Information | |
|---|---|
| Course attendance is strongly recommended. Students must be prepared to spend a fair amount of time in the study outside of lessons. For a satisfactory preparation students need to spend an average of one hour of study time for each hour spent in class. Those who decide do develop the optional project should spend about 80 hours for developing it. Course materials will be available for download from the departmental e-learning platform http://elearning.informatica.unisa.it/el-platform/ Contacts Prof. Loredana Caruccio lcaruccio@unisa.it |
BETA VERSION Data source ESSE3
