(2019, June), Introducing machine learning concepts using hands-on Android-based exercises  Paper presented at 2019 ASEE Annual Conference & Exposition , Tampa, Florida. https://peer.asee.org/33014

\bibitem{asee_peer_33014}
   (2019, June), \emph{Introducing machine learning concepts using hands-on Android-based exercises}
  Paper presented at 2019 ASEE Annual Conference \& Exposition , Tampa, Florida.
  https://peer.asee.org/33014

.  "Introducing machine learning concepts using hands-on Android-based exercises".  2019 ASEE Annual Conference & Exposition , Tampa, Florida, 2019, June.  ASEE Conferences, 2019.  https://peer.asee.org/33014 Internet.  21 Apr, 2024

\bibitem{asee_peer_33014}
  .
  "Introducing machine learning concepts using hands-on Android-based exercises".
  \emph{2019 ASEE Annual Conference \& Exposition , Tampa, Florida, 2019, June}.
  ASEE Conferences, 2019.
  https://peer.asee.org/33014 Internet.  21 Apr, 2024

@INPROCEEDINGS{asee_peer_33014
author = ""
title = "Introducing machine learning concepts using hands-on Android-based exercises"
booktitle = "2019 ASEE Annual Conference \& Exposition "
year = "2019"
month = "June"
address = "Tampa, Florida"
publisher = "ASEE Conferences"
note = {https://peer.asee.org/33014}
}

TY  - CPAPER
AB  - Teaching machine learning involves introducing students to complex concepts in statistics and optimization. In order for students to better grasp concepts in machine learning, we provide them with hands-on exercises. We provide the students with two different sets of activities, both of which involve data collection, modeling, training, and testing, as applied to machine learning algorithms. These types of immersive experiences will provide students with exposure to the different stages of the practical uses of machine learning, expose them to different types of machine learning algorithms, and give them exposure to how modeling affects performance. The data collection apparatus is based on apps developed for the Android platform. 

In the first set of experiments, students use an app that collects touch and swipe data. Each student is asked to perform several activities such as tracing over a shape, typing a block of text, and selecting a specific picture from a gallery. The touch data while the student interacts with the device are recorded and stored separately for each student. Using this labeled data, students then select a machine learning algorithm, train a model, and then perform testing, with the goal being to distinguish users based on touch data. 

In the second experiment, an audio signal is generated from an app, and the reflection is recorded, to give us an “audio profile” of the environment. Different rooms are selected and in each room, several audio profiles are recorded. Students then train and test machine learning models that distinguish between the rooms based on the recorded audio profiles. 

Performance results are presented to students in each case. Students will also be able to select from several machine learning algorithms, and manually modify trained models and study the effects on performance. Concepts that can be covered include applications of machine learning and how machine learning provides the basis for decision making and artificial intelligence. 

We will hold workshops to evaluate the effectiveness of this approach. The assessment results will be presented at the conference. 

This work is supported through an NSF IUSE (DUE) award. 

CY  - Tampa, Florida
DA  - 2019/06/15
PB  - ASEE Conferences
TI  - Introducing machine learning concepts using hands-on Android-based exercises
UR  - https://peer.asee.org/33014
ER  -