I am a Ph.D. candidate in computer science in the school of Electrical Engineering and Computer Science at Washington State University. I work as a teaching and research assistant under the supervision of Dr.Hassan Ghasemzadeh in Embedded and Pervasive Systems Laboratory. My research focuses on personalized autonomous learning for wearable and mobile health monitoring. I have experience with machine learning, data science, and software development.
Developed machine learning and data science methods for analyzing and predicting from time-series data
Developed Android data collection mobile applications◦Served as a mentor to several undergraduate and graduate-level students
TA: Computer Architecture, Programming in Python, Algorithm Design, and Microprocessors
Developed deep learning models for lying posture tracking using inertial sensor data
Programming, testing, and debugging banking software solutions using C-Sharp.Net and SQL.
We introduce a power-aware sensor localization approach that allows users to wear the sensors on different body locations without the need for adhering to a speciﬁc installation protocol. Furthermore, we propose a novel transductive transfer learning approach, which gives end-users the ability to add new sensors to the network without the need for collecting new training data. This is accomplished by transferring the knowledge of already trained sensors to the untrained sensors in real-time.
We introduce a filter-based step counting algorithm that is reliable against various walking speeds and intensities. ParaLabel aims to learn the algorithm parameters autonomously in new domains (e.g., in new users) without a need for collecting sensor data and manually tuning the algorithm parameters. We formulate this problem as a transfer learning problem where parameters of the algorithm in a new domain are mapped onto a parameter bank containing previously fine-tuned parameters in a different domain.
In this study, we implement two models (feature-based and deep learning models) to detect three main lying postures (lying supine, prone, and on the sides) from a single tri-axial accelerometer sensor. The feature-based model is an ensemble tree classifier which is built on 48 time-domain features. The deep learning model is a long short-term memory (LSTM) classifier that is trained on raw accelerometer data. In this study, we 1) compare performance of feature-based and deep learning models for lying posture detection, 2) identify the optimal sensor wearing site (from nine different body locations including the chest, right and left thigh, ankle, arm, and wrist) for accurate lying posture detection using a single accelerometer, and 3) identify the optimal set of time-domain features in feature-based model for accurate lying posture detection.
J.1. A Reliable and Reconfigurable Signal Processing Framework for Estimation of Metabolic Equivalent of Task in Wearable Sensors, Parastoo Alinia, Ramyar Saeedi, Ramin Fallahzadeh, Ali Rokni, Hassan Ghasemzadeh, IEEEJournal of Selected Topics in Signal Processing
J.1. Multi-sensor fusion in body sensor networks: State-of-the-art and research challenges, Raffaele Gravina, Parastoo Alinia, Hassan Ghasemzadeh, Giancarlo Fortino, Information Fusion
J.2. How Accurate Is Your Activity Tracker? A Comparative Study of Step Counts in Low-Intensity Physical Activities, Parastoo Alinia Chris Cain, Ramin Fallahzadeh, Armin Shahrokni, Diane Cook, Hassan Ghasemzadeh, JMIR mHealth, and uHealth
J.3. [Under Review] ParaLabel: An Autonomous Parameter Learning Approach for Cross-Domain Step Count-ing in Wearable Sensors, Parastoo Alinia Ramin Fallahzadeh, Christopher Connolly, and Hassan Ghasemzadeh, ACMTransactions on Internet of Things (TIOT)
J.4. [Under Review] Acceleration-based Lying Posture Tracking: Traditional Classifier VS. Deep Learning, Paratoo Alinia, Ali Samadani, Mladen Milosevic, Hassan Ghasemzadeh, and Saman Parvaneh, Journal of Pervasive and Mobile Computing
C.1. Impact of Sensor Misplacement on Calculating Metabolic Equivalent of Task with Wearables, Parastoo Alinia Ramyar Saeedi, Bobak Mortazavi, Ali Rokni, Hassan Ghasemzadeh, In Wearable and Implantable BodySensor Networks (BSN), IEEE 12th International Conference on IEEE
C.2. An Energy Efficient Computational Model for Uncertainty Management in Dynamically Changing Networked Wearables, Ramyar Saeedi, Ramin Fallahzadeh, Parastoo Alinia, Hassan Ghasemzadeh, In Proceedings of the 2016 International Symposium on Low Power Electronics and Design
C.3.Learn-on-the-Go: Autonomous Cross-Subject Context Learning for Internet-of-Things Applications HG Ramin Fallahzadeh, Parastoo Alinia, The 36th IEEE/ACM International Conference On Computer-AidedDesign (ICCAD)2018
A.1. [Abstract] Edema Monitoring System, ACM Conference on Wireless Health (WH), October 14–16, 2015, National Institutes of Health, Bethesda, MD, Parastoo Alinia , Ramin Fallahzadeh, Ramyar Saeedi, Hassan Ghasemzadeh
A.2. [Patent--Pending] Efficient Lying Posture Tracking with Philips Research North America, Parastoo Alinia and Saman Parvaneh
A.3. [Book Chapter--Pending] Reliable and Power-Efficient Machine Learning in Wearable Sensors, Parastoo Alinia andHassan Ghasemzadeh Fog Computing: Theory and Practice, WILEY