MDMichael Domaratzki

Michael Domaratzki

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Michael Domaratzki has a total of 101 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    LSTM-attention-guided graph neural networks for integrated genotype–Environment modeling in maize yield prediction

    by Amir Morshedian, Mike Domaratzki

    This paper presents a deep-learning framework that combines an LSTM, a graph neural network (GNN), and transformer-style attention to model genotype–environment (G×E) effects for maize yield prediction. Weather data for a growing season is summarized using LSTM and encoded into a 21-dimensional embedding that is used as the environment node feature; 437,214 SNPs are summarized into 548 principal components that instantiate genotype nodes. Multi-head attention dynamically weights the edges during message passing. Three architectures are compared: A (fully bipartite graph), B (A with intra-set top- k similarity within genotype and within environment), and C (B with a single learnable supernode readout that attends over all nodes after message passing). The joint representations feed a compact MLP for yield prediction. Using a forward-time split (2014–2021 train; 2022 test with unseen genotypes and unseen environments), performance improves monotonically from A to C: A (RMSE 2.7749, PCC 0.4115, R 2 0.1693), B (2.3683, 0.6622, 0.4385), C (2.2120, 0.6945, 0.4823). Compared to A, C has a reduction in RMSE by 0.5629 (∼20.3%) and an increase in PCC by 0.283 (∼68.8%), indicating that global, content-adaptive aggregation promotes local G×E propagation. Performance of proposed approach remains consistent regardless of the number of genotypes per environment and has strong performance under variable or unbalanced genotype sampling expression across environments. The proposed approach is compared with methods from the Global G×E Prediction Competition and show that two of three architectures improve predictive performance, with the best architecture achieving a lower RMSE (2.2120) and a higher Pearson correlation (0.6945) than the competition-winning model.


  • Open Access

    Machine learning of intraoperative variables to test feasibility of multivariable prediction modelling for postoperative complications in thoracic surgery: a prospective cohort study

    by Biniam Kidane, Atif Ul Aftab, Eagan J. Peters, Sadeesh Srinathan, Gordon Buduhan, Lawrence Tan, Emma Poole, Michael Domaratzki

    Background: Among patients undergoing thoracic surgery, the impact of intraoperative variables on postoperative complications is unclear. Because patients receiving one-lung ventilation (OLV) experience further unique intraoperative stressors, a knowledge gap exists around the impact of intraoperative predictor variables that may not be well-accounted for in existing risk prediction models. The objectives of this study were therefore to (I) assess the feasibility of measuring intraoperative variables using modern machine learning techniques; (II) determine if machine learning of intraoperative parameters predicts postoperative complications; and (III) compare model performance of machine learning against a set of known preoperative predictors. Methods: A prospective cohort study was performed of consecutive patients undergoing thoracic surgery with OLV at a single Canadian centre. Intraoperative ventilatory and hemodynamic data were captured. Machine learning was used to predict postoperative complications based on: (I) preoperative variables only; (II) intraoperative variables only; and (III) combined data from both preoperative and intraoperative variables. Machine learning classification algorithms using random forests, support vector machines, and logistic regression models were compared. These models were analyzed with and without synthetic minority over-sampling technique (SMOTE). All models used the same split of training and test data. Results: Of 121 surgeries, there were 54 (44.6%) sublobar resections, 51 (42.2%) lobectomies, 3 (2.5%) pneumonectomies, and 13 (10.7%) non-pulmonary surgeries. 100 (82.6%) surgeries were minimally invasive. Mean operative time was 122.3±76.4 minutes. Postoperative complications occurred in 35 (28.9%) patients. Support vector machine classification algorithm using SMOTE of intraoperative data alone predicted complications with the highest F1 score of 0.8298. The most important parameters driving prediction accuracy were (I) duration of operative time; (II) mean and maximum heart rate; and (III) fraction of inspired oxygen. Using preoperative data alone and adding preoperative data to intraoperative data did not improve F1 score (max F1 score =0.6643 and 0.7647 respectively). Conclusions: Machine learning of intraoperative data may be feasible to predict postoperative complications. Therefore, intraoperative data could be used to augment existing risk prediction models. However, this study is hypothesis-generating only. Analysis of larger scale samples focused on specific complications by type are required to improve predictions.

  • Open Access

    Regression augmentation with data-driven segmentation

    by Shayan Alahyari, Shiva Mehdipour Ghobadlou, Mike Domaratzki


  • Open Access

    Obscured-ensemble models for genomic prediction

    by Rounak Saha, Amir Morshedian, Jia Sun, Robert Duncan, Mike Domaratzki

    Genomic Prediction (GP) uses dense whole-genome marker sets from lines of a crop to predict agronomic traits for untested genotypes. In recent years, deep learning (DL) approaches for genomic prediction have demonstrated state-of-the-art results. However, substantial variation exists in DL outcomes for GP as the success of DL is dependent on the architecture of the model used, as well as the amount of data available and the population structure of the individuals in the training set. In this paper, we consider an obscured model for GP, where the model is not provided with genomic content. The obscured model was intended to evaluate the possibility of so-called shortcut learning in GP.We conclude that we can perform GP using the obscured model with only 20% of the obscured markers from each reference genotype. This selective feature usage significantly enhances the efficiency of our model without compromising accuracy. By eliminating markers, we demonstrate that the model is not relying on linkage to perform shortcut learning. Further, we consider a deep learning ensemble method for genomic prediction based on the obscured model. The ensemble model we develop here shows success as a method for GP by using the similarity to each of the elements of a training set of genotypes, as well as the performance of the genotypes. We evaluate the obscured ensemble model for GP. We demonstrate that the obscured ensemble model is successful even with a limited number of genotypes used for prediction. Further, random selection of a subset of genotypes is sufficient to ensure successful performance.


  • Open Access

    LSTM-Attention-Guided Graph Neural Networks for Integrated Genotype–Environment Modeling in Maize Yield Prediction

    by Amir Morshedian, Mike Domaratzki

    We present a deep-learning framework that combines an LSTM, a graph neural network (GNN), and transformer-style attention to model genotype–environment (G × E) effects for maize yield prediction. Weather data for a growing season is summarized using LSTM and encoded into a 21-dimensional embedding that is used as the environment node feature; 437,214 SNPs are summarized into 548 principal components that instantiate genotype nodes. Multi-head attention dynamically weights the edges during message passing. We compare three architectures: A (fully bipartite graph), B (A with intra-set top- k similarity within genotype and within environment), and C (B with a single learnable supernode readout that attends over all nodes after message passing). The joint representations feed a compact MLP for yield prediction. Using a forward-time split (2014–2021 train; 2022 test with unseen genotypes and unseen environments), performance improves monotonically from A to C: A (RMSE 2.7749, PCC 0.4115, R 2 0.1693), B (2.3683, 0.6622, 0.4385), C (2.2120, 0.6945, 0.4823). Compared to A, C has a reduction in RMSE by 0.5629 (∼20.3%) and an increase in PCC by 0.283 (∼68.8%), indicating that global, content-adaptive aggregation promotes local G × E propagation. Performance of our approach remains consistent regardless of the number of genotypes per environment and has strong performance under variable or unbalanced genotype sampling expression across environments.

  • Open Access

    Regression Augmentation With Data-Driven Segmentation

    by Shayan Alahyari, Shiva Mehdipour Ghobadlou, Mike Domaratzki

    Imbalanced regression arises when the target distribution is skewed, causing models to focus on dense regions and struggle with underrepresented (minority) samples. Despite its relevance across many applications, few methods have been designed specifically for this challenge. Existing approaches often rely on fixed, ad hoc thresholds to label samples as rare or common, overlooking the continuous complexity of the joint feature-target space and fail to represent the true underlying rare regions. To address these limitations, we propose a fully data-driven GAN-based augmentation framework that uses Mahalanobis-Gaussian Mixture Modeling (GMM) to automatically identify minority samples and employs deterministic nearest-neighbour matching to enrich sparse regions. Rather than preset thresholds, our method lets the data determine which observations are truly rare. Evaluation on 32 benchmark imbalanced regression datasets demonstrates that our approach consistently outperforms state-of-the-art data augmentation methods.

  • Open Access

    Minimizing features while maintaining performance in data classification problems

    by Surani Matharaarachchi, Mike Domaratzki, Saman Muthukumarana

    High dimensional classification problems have gained increasing attention in machine learning, and feature selection has become essential in executing machine learning algorithms. In general, most feature selection methods compare the scores of several feature subsets and select the one that gives the maximum score. There may be other selections of a lower number of features with a lower score, yet the difference is negligible. This article proposes and applies an extended version of such feature selection methods, which selects a smaller feature subset with similar performance to the original subset under a pre-defined threshold. It further validates the suggested extended version of the Principal Component Loading Feature Selection (PCLFS-ext) results by simulating data for several practical scenarios with different numbers of features and different imbalance rates on several classification methods. Our simulated results show that the proposed method outperforms the original PCLFS and existing Recursive Feature Elimination (RFE) by giving reasonable feature reduction on various data sets, which is important in some applications.


  • Open Access

    Discovering Long COVID Symptom Patterns: Association Rule Mining and Sentiment Analysis in Social Media Tweets

    by Surani Matharaarachchi, Mike Domaratzki, Alan Katz, Saman Muthukumarana

    Background: The COVID-19 pandemic is a substantial public health crisis that negatively affects human health and well-being. As a result of being infected with the coronavirus, patients can experience long-term health effects called long COVID syndrome. Multiple symptoms characterize this syndrome, and it is crucial to identify these symptoms as they may negatively impact patients’ day-to-day lives. Breathlessness, fatigue, and brain fog are the 3 most common continuing and debilitating symptoms that patients with long COVID have reported, often months after the onset of COVID-19. Objective: This study aimed to understand the patterns and behavior of long COVID symptoms reported by patients on the Twitter social media platform, which is vital to improving our understanding of long COVID. Methods: Long COVID–related Twitter data were collected from May 1, 2020, to December 31, 2021. We used association rule mining techniques to identify frequent symptoms and establish relationships between symptoms among patients with long COVID in Twitter social media discussions. The highest confidence level–based detection was used to determine the most significant rules with 10% minimum confidence and 0.01% minimum support with a positive lift. Results: Among the 30,327 tweets included in our study, the most frequent symptoms were brain fog (n=7812, 25.8%), fatigue (n=5284, 17.4%), breathing/lung issues (n=4750, 15.7%), heart issues (n=2900, 9.6%), flu symptoms (n=2824, 9.3%), depression (n=2256, 7.4%) and general pains (n=1786, 5.9%). Loss of smell and taste, cold, cough, chest pain, fever, headache, and arm pain emerged in 1.6% (n=474) to 5.3% (n=1616) of patients with long COVID. Furthermore, the highest confidence level–based detection successfully demonstrates the potential of association analysis and the Apriori algorithm to establish patterns to explore 57 meaningful relationship rules among long COVID symptoms. The strongest relationship revealed that patients with lung/breathing problems and loss of taste are likely to have a loss of smell with 77% confidence. Conclusions: There are very active social media discussions that could support the growing understanding of COVID-19 and its long-term impact. These discussions enable a potential field of research to analyze the behavior of long COVID syndrome. Exploratory data analysis using natural language processing methods revealed the symptoms and medical conditions related to long COVID discussions on the Twitter social media platform. Using Apriori algorithm–based association rules, we determined interesting and meaningful relationships between symptoms.