Engineering an IoT-based data collection system for predictive Machine Learning algorithm use
Kho Caayon, Arthur (2018)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2018120620390
https://urn.fi/URN:NBN:fi:amk-2018120620390
Tiivistelmä
Every physical or virtual activity can be turned into data. And the challenge is two-fold. First, in identifying and capturing events as they happen. Second, in converting them into a manipulable form.
This project is an attempt in tackling the challenge posed by a specific scenario. That of how to come up with ways and means of identifying and capturing the artifact of the passive interaction between a group of students and a computerized “intelligent system” where the context for the passive interaction is inside a classroom.
The solution is through the use of an array of electronic sensors that is controlled by a smart system. The idea is to use the sensors to capture events of interest. The smart system then converts those sensor readings into something that a Machine Learning algorithm can process.
A proof-of-concept was produced. The scale and scope of the solution was deliberately constrained to that of only the hardware side.
This project is an attempt in tackling the challenge posed by a specific scenario. That of how to come up with ways and means of identifying and capturing the artifact of the passive interaction between a group of students and a computerized “intelligent system” where the context for the passive interaction is inside a classroom.
The solution is through the use of an array of electronic sensors that is controlled by a smart system. The idea is to use the sensors to capture events of interest. The smart system then converts those sensor readings into something that a Machine Learning algorithm can process.
A proof-of-concept was produced. The scale and scope of the solution was deliberately constrained to that of only the hardware side.