This is the home for open code, data and research publications for MLGIG (Machine Learning Research Group of Dr Georgiana Ifrim). Below is a list of our public repositories organised mainly into Data Challenges, Time Series Classification (algorithms, applications), Time Series Explanation, Sequence Learning and Text Mining. A recent presentation provides a short overview of some of these projects.
Where we apply our knowledge of machine learning modeling and time series modeling:
- 2024 VistaMilk-challenge-lactose-prediction-regression (1st place)
- 2022 VistaMilk-challenge-diet-classification (1st place)
- 2021 VistaMilk-spectroscopy-challenge-predict-milk-traits-regression (2nd place)
- 2014 SNOW Data Challenge (1st prize): Twitter Topics (1st place)
- MRSQM: Multiple Representation Sequence Miner
- MRSEQL: Multiple Representation Sequence Learner
- Fast Channel Selection for Multivariate Time Series Classification
- Back to Basics: A Sanity Check on Modern Time Series Classification Algorithms
- An Examination of the State-of-the-Art for Multivariate Time Series Classification
- 2024 Accurate and Efficient Real-World Fall Detection Using Time Series Techniques
- 2023 An Examination of Wearable Sensors versus Video Data Capture for Human Exercise Classification
- 2022 Fast and Robust Video-Based Exercise Classification via Body Pose Tracking and Scalable Multivariate Time Series Classifiers
- 2021 Interpretable Classification of Human Exercise Videos through Pose Estimation and Multivariate Time Series Analysis
- 2024 tsCaptum - A library for univariate and multivariate time series explanation
- 2024 Improving the Evaluation and Actionability of Explanation Methods for Multivariate Time Series Classification
- 2024 AMEE: Robust explainer recommendation for time series classification
- 2023 Evaluating-Explanation-Methods-for-MTSC