Marija Lazaroska Email & Phone Number

Switzerland

Gland, Vaud, Switzerland

Lausanne, Vaud, Switzerland

I worked on my Master's Thesis on the topic of Explainable AI. My responsabilities were:- Research and implementation of state-of-the-art explainability methods(SHAP, LIME, TabNet) for creating explanations of black-box models.- Development of a new explainability method using the Attention Mechanism to explain deep learning models trained with tabular data—a full pipeline implementation in PyTorch.- Participation in different activities to promote knowledge sharing across team… Show more I worked on my Master's Thesis on the topic of Explainable AI. My responsabilities were:- Research and implementation of state-of-the-art explainability methods(SHAP, LIME, TabNet) for creating explanations of black-box models.- Development of a new explainability method using the Attention Mechanism to explain deep learning models trained with tabular data—a full pipeline implementation in PyTorch.- Participation in different activities to promote knowledge sharing across team members and sharing the work with other colleagues in the company through different presentations every week. Show less

Lausanne, Vaud, Switzerland

As a part-time student researcher, I worked on two projects from the LHTC lab, compromising the innovation of a non-invasive method for obtaining cardiac output(CO) and aortic systolic blood pressure(aSBP). I successfully provided a method using CNN trained on brachial waveforms, giving an R-squared score of 0.92. My responsabilities were:- Implementation of CNN models that predict the aortic systolic blood pressure (aSBP) and the cardiac output (CO) from the brachial pressure… Show more As a part-time student researcher, I worked on two projects from the LHTC lab, compromising the innovation of a non-invasive method for obtaining cardiac output(CO) and aortic systolic blood pressure(aSBP). I successfully provided a method using CNN trained on brachial waveforms, giving an R-squared score of 0.92. My responsabilities were:- Implementation of CNN models that predict the aortic systolic blood pressure (aSBP) and the cardiac output (CO) from the brachial pressure waveforms.- Development of an end-to-end pipeline in PyTorch, comprising pre-processing of time series and training/testing CNN models.- Testing on synthetic data and performing adjustments to adapt it to real data in an environment with reduced data samples. Show less

Lausanne, Vaud, Switzerland

I worked on the detection of cough patterns in time-series data from physiological measurements coming from bed sensors. My responsibilities were:- Creating a dataset from time-series data using Pandas and TFRecord to be used for analytics and machine learning tasks.- Implementation of a self-supervised model (i.e., VICReg, VIBCReg) using PyTorch and package the code to integrate into the company’s code stack.- Develop visualisation methods (UMAP, PCA) using JavaScript and… Show more I worked on the detection of cough patterns in time-series data from physiological measurements coming from bed sensors. My responsibilities were:- Creating a dataset from time-series data using Pandas and TFRecord to be used for analytics and machine learning tasks.- Implementation of a self-supervised model (i.e., VICReg, VIBCReg) using PyTorch and package the code to integrate into the company’s code stack.- Develop visualisation methods (UMAP, PCA) using JavaScript and Plotly to visualise and evaluate the results from the machine learning models. Show less

Master Of Computer Science

Bachelor'S Degree, Computer Science

Second and third year of Bachelor studies.

Computer Science

First year of Bachelor studies. Finished 2nd place.