Publications

2024

1. Adaptive Learning Rates for Gradient Boosting Machines

   Christopher Wang, Zheng Wang , Yunfei Ouyang , and Behrouz Haji Soleimani

   Proceedings of the Canadian Conference on Artificial Intelligence, May 2024

   Abs Bib HTML

   Gradient Boosting Machines (GBM) is a widely applicable machine learning algorithm that has demonstrated top performance in a variety of fields. In this paper, we explore the potential of adaptive learning rates to achieve accelerated convergence in GBMs. We introduce a novel boosting algorithm called Delta-Bar-Delta (DBD) Boosting that leverages insights from the steepest-descent algorithm of the same name. We show improved performance over the baseline GBM model through a series of experiments. We also show that our proposed DBD boosting algorithm can be conveniently combined with other optimization improvements, such as momentum and Nesterov’s Accelerated Gradient. We perform hyperparameter tuning and evaluate our algorithm on series of classification and regression tasks. Our findings demonstrate empirically improved convergence rate compared to existing approaches. Furthermore, we observe and discuss intriguing behaviors related to adaptive learning rates within the context of GBMs, highlighting the intricate dynamics of our proposed method. This research contributes to the ongoing advancement of gradient boosting techniques in machine learning, offering new perspectives and tools for improved convergence and faster training.

   @article{Wang2024Adaptive,
     author = {Wang, Christopher and Wang, Zheng and Ouyang, Yunfei and Soleimani, Behrouz Haji},
     journal = {Proceedings of the Canadian Conference on Artificial Intelligence},
     year = {2024},
     month = may,
     publisher = {Canadian Artificial Intelligence Association (CAIAC)},
     title = {Adaptive {Learning} {Rates} for {Gradient} {Boosting} {Machines}},
   }

2022

1. Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI)

   Kevin Dick , Daniel G. Kyrollos , Eric D. Cosoreanu , Joseph Dooley , Joshua S. Fryer , Shaun M. Gordon , Nikhil Kharbanda , Martin Klamrowski , Patrick N. L. LaCasse , Thomas F. Leung , Muneeb A. Nasir , Chang Qiu , Aisha S. Robinson , Derek Shao , Boyan R. Siromahov , Evening Starlight , Christophe Tran , Christopher Wang, Yu-Kai Yang , and James R. Green

   Scientific Reports, Aug 2022

   Abs Bib HTML

   The identification of novel drug-target interactions (DTI) is critical to drug discovery and drug repurposing to address contemporary medical and public health challenges presented by emergent diseases. Historically, computational methods have framed DTI prediction as a binary classification problem (indicating whether or not a drug physically interacts with a given protein target); however, framing the problem instead as a regression-based prediction of the physiochemical binding affinity is more meaningful. With growing databases of experimentally derived drug-target interactions (e.g. Davis, Binding-DB, and Kiba), deep learning-based DTI predictors can be effectively leveraged to achieve state-of-the-art (SOTA) performance. In this work, we formulated a DTI competition as part of the coursework for a senior undergraduate machine learning course and challenged students to generate component DTI models that might surpass SOTA models and effectively combine these component models as part of a meta-model using the Reciprocal Perspective (RP) multi-view learning framework. Following 6 weeks of concerted effort, 28 student-produced component deep-learning DTI models were leveraged in this work to produce a new SOTA RP-DTI model, denoted the Meta Undergraduate Student DTI (MUSDTI) model. Through a series of experiments we demonstrate that (1) RP can considerably improve SOTA DTI prediction, (2) our new double-cold experimental design is more appropriate for emergent DTI challenges, (3) that our novel MUSDTI meta-model outperforms SOTA models, (4) that RP can improve upon individual models as an ensembling method, and finally, (5) RP can be utilized for low computation transfer learning. This work introduces a number of important revelations for the field of DTI prediction and sequence-based, pairwise prediction in general.

   @article{musdti,
     author = {Dick, Kevin and Kyrollos, Daniel G. and Cosoreanu, Eric D. and Dooley, Joseph and Fryer, Joshua S. and Gordon, Shaun M. and Kharbanda, Nikhil and Klamrowski, Martin and LaCasse, Patrick N. L. and Leung, Thomas F. and Nasir, Muneeb A. and Qiu, Chang and Robinson, Aisha S. and Shao, Derek and Siromahov, Boyan R. and Starlight, Evening and Tran, Christophe and Wang, Christopher and Yang, Yu-Kai and Green, James R.},
     title = {Reciprocal perspective as a super learner improves drug-target interaction prediction (MUSDTI)},
     journal = {Scientific Reports},
     year = {2022},
     month = aug,
     day = {02},
     volume = {12},
     number = {1},
     pages = {13237},
     issn = {2045-2322},
     doi = {10.1038/s41598-022-16493-9},
   }

2. Canadian Jobs amid a Pandemic: Examining the Relationship between Professional Industry and Salary to Regional Key Performance Indicators

   Rahul Anilkumar , Benjamin Melone , Michael Patsula , Christophe Tran , Christopher Wang, Kevin Dick , Hoda Khalil , and Gabriel Wainer

   In IEEE COMPSAC 2022 , Jun 2022

   Abs Bib HTML

   The COVID-19 pandemic has contributed to un-precedented rates of unemployment and greater uncertainty in the job market. There is a growing need for data-driven tools and analyses to better inform the public on trends within the job market. In particular, obtaining a “snapshot” of available employment opportunities mid-pandemic promises insights to inform policy and support retraining programs. In this work, we combine data scraped from the Canadian Job Bank and Numbeo globally crowd-sourced repository to explore the relationship between job postings during a global pandemic and Key Performance Indicators (e.g. quality of life [QOL] index, cost of living) for major cities across Canada. This analysis aims to help Canadians make informed career decisions, collect a “snapshot” of the Canadian employment opportunities amid a pandemic, and inform job seekers in identifying the correct fit between the desired lifestyle of a city and their career. We collected a new high-quality dataset of job postings from jobbank.gc.ca obtained with the use of ethical web scraping and performed exploratory data analysis on this dataset to identify job opportunity trends. When optimizing for average salary of job openings with QOL, affordability, cost of living, and traffic indices, it was found that Edmonton, AB consistently scores higher than the mean, and is therefore an attractive place to move. Furthermore, we identified optimal provinces to relocate to with respect to individual skill levels. It was determined that Ajax, Marathon, and Chapleau, ON are each attractive cities for IT professionals, construction workers, and healthcare workers respectively when maximizing average salary. Finally, we publicly release our scraped dataset as a mid-pandemic snapshot of Canadian employment opportunities and present a public web application that provides an interactive visual interface that summarizes our findings for the general public and the broader research community.

   @inproceedings{kpi,
     author = {Anilkumar, Rahul and Melone, Benjamin and Patsula, Michael and Tran, Christophe and Wang, Christopher and Dick, Kevin and Khalil, Hoda and Wainer, Gabriel},
     year = {2022},
     month = jun,
     pages = {235-240},
     title = {Canadian Jobs amid a Pandemic: Examining the Relationship between Professional Industry and Salary to Regional Key Performance Indicators},
     doi = {10.1109/COMPSAC54236.2022.00041},
     booktitle = {IEEE COMPSAC 2022},
   }

2019

1. An ensemble of U-Net architecture variants for left atrial segmentation

   Christopher Wang, Eranga Ukwatta , Martin Rajchl , and Adrian Chan

   In SPIE Medical Imaging 2019 , Mar 2019

   Abs Bib HTML

   Segmentation of the left atrium and proximal pulmonary veins is an important clinical step for diagnosis of atrial fibrillation. However, the automatic segmentation of the left atrium from late gadolinium-enhanced magnetic resonance (LGE-MRI) images remains a challenging task due to differences in acquisition and large variability between individuals. Deep learning has shown to outperform traditional methodologies for segmentation in numerous tasks. A popular deep learning architecture for segmentation is the U-Net, which has shown promising results biomedical segmentation problems. Many newer network architectures have been proposed that leverage the base U-Net architecture such as attention U-Net, dense U-Net and residual U-Net. These models incorporate updated encoder blocks into the U-Net architecture to incrementally improve performance over the base U-Net. Currently, there is no comprehensive evaluation of performance between these models. In this study we (1) explore approaches for the segmentation of the left atrium based on different- Net architectures. (2) We compare and evaluate these on the STACOM 2018 Atrial Segmentation Challenge dataset and (3) ensemble these models to improve overall segmentation by reducing the internal variance between models and architectures. (4) Lastly, we define and build upon a U-Net framework to simplify development of novel U-Net inspired architectures. Our ensemble achieves a mean Dice similarity coefficient (DSC) of 92.1 ± 2.0% on a test set of twenty 3D LGE-MRI images, outperforming other fully automatic segmentation methodologies.

   @inproceedings{unet,
     author = {Wang, Christopher and Ukwatta, Eranga and Rajchl, Martin and Chan, Adrian},
     year = {2019},
     month = mar,
     pages = {21},
     title = {An ensemble of U-Net architecture variants for left atrial segmentation},
     doi = {10.1117/12.2512905},
     booktitle = {SPIE Medical Imaging 2019},
   }