Time series classification for food safety using advanced machine learning
A cutting-edge solution for classifying biosensor outputs using machine learning and deep learning techniques to enhance food safety. This approach involves converting time series data to images and applying multimodal fusion for improved accuracy.
Overview
This innovative solution leverages advanced machine learning and deep learning algorithms to classify biosensor outputs, crucial for detecting food safety targets. By transforming time series data into images and applying multimodal fusion techniques, this approach aims to match or exceed the accuracy of existing gold standards in the industry. The solution is poised to improve the detection and analysis of food safety markers by up to 5% beyond current methods.
Technical specifications
- Data Transformation: Converts time series data from biosensors into 2D images, facilitating the use of sophisticated image analysis techniques.
- Multimodal Fusion: Integrates multiple data sources and modalities to enhance classification accuracy and reliability.
- Model Development: Initial models will be developed and validated with limited data, followed by iterative improvements and evaluations on larger datasets.
- Prototype Software: A user-friendly software prototype will be developed to enable partners to easily deploy and utilize the classification system.
Technology readiness level
This technology is currently at TRL 2, indicating that its principles have been formulated, and it is in the early stages of development. The project is set to progress through rigorous validation and improvement phases over a two-year period, culminating in a deployable prototype software for industry partners.
About Toronto Metropolitan University
Toronto Metropolitan University is a comprehensive public research university in the heart of downtown Toronto, recognized for experiential, career‑integrated education and a diverse, industry‑connected student body. Industry access is embedded through co‑located and community‑based platforms, including the DMZ startup incubator and a university‑wide Zone Learning network that connects companies with faculty and student innovators. A dedicated Centre for Urban Innovation hosts collaborative labs and pilot spaces, while the Brampton‑based School of Medicine—affiliated with William Osler Health System—creates direct pathways for clinical training and partnership; the Rogers Cybersecure Catalyst extends capacity in workforce development and applied solutions. Research is supported by competitive federal and provincial funding, including NSERC, CIHR, SSHRC and the Canada Foundation for Innovation. A formal commercialization framework and IP policy provide structured support for protection, licensing and venture creation.
delivered to your inbox.