Here is a list of data science projects which I conducted recently.
Developing Data Quality Common Tools for the European Space Agency's Euclid satellite.
Developing Tools required for the quality control of the observational images and all data produced by the Euclid Satellite
The project involved developing data quality common tools (DQCTs) in Python 3 and C++ for the European Space Agency's Euclid satellite to ensure the quality control of observational images and all data produced. I contributed to designing and developing algorithms, defining parameters, and providing coding tutorials and wiki pages, including relevant practical examples.
Lung segmentation from Chest X-Ray dataset and Quality Control of
The project involved developing an open-source Artificial Intelligence tool that combines chest imaging, clinical, and laboratory data to support the diagnosis, triaging, and treatment planning for COVID-19.
Markov Chain Monte Carlo (MCMC) sampling and implementation
Modeling scaling relation of galaxies and groups
Scaling relations describe strong trends observed between important physical properties (such as mass, size, luminosity, and colors) of galaxies. Here I measure the Lx − σv relation of groups of galaxies (see Gozaliasl et al. 2020).
Galaxy Group Detection and Membership Determination
Galaxy cluster membership selection
With Dr. Francesco Montanari, we developed a code (GitHub link) to detect clusters and group membership in observation and the N-body numerical or hydrodynamical simulations.
Sensor Data for IoT Predictive Analytics
The individual has extensive experience in time-series data analytics, specifically using sensor data. Recently, they conducted a project on road weather forecasting, which involved developing a state-of-the-art service that can predict current and near-future driving conditions. The system used data from standard professional sensors and road weather stations along the route to generate accurate weather forecasts.
The individual played a significant role in the project, designing and implementing the data analytics algorithms that processed the sensor data and generated the weather forecasts. They were also responsible for testing and validating the system to ensure its accuracy and reliability.
The road weather forecasting system has important public safety and transportation infrastructure management implications. Accurate weather forecasts can help drivers make informed travel plans and improve road maintenance and planning for winter conditions.
Prediction of APP installs probability for ad impressions.
In this project, I conducted research on game companies' use of deep learning models for real-time ad bidding (OpenRTB) at various ad exchanges. I utilized different machine learning algorithms and neural network models to predict install probabilities for ad impressions, alongside other factors such as cost per installation, to estimate the optimal bid value for an ad impression. Through this project, I gained experience in developing predictive models and knowledge in real-time ad bidding. This project's results had significant contributions and implications for the gaming industry's advertising strategies and demonstrate their proficiency in machine learning and neural network modeling.
Automated Detection, Classification, and Counting of Fish in Fish Passages With Deep Learning
The project aimed to develop a system to classify, track, and count fish as they pass in front of a camera located within a river. The system utilized computer vision techniques to analyze the video feed from the camera and identify different fish species based on their unique visual characteristics. The system could also track individual fish moving through the camera's field of view, allowing for accurate counting of fish populations.
I was involved in the project as the project owner's role in developing and implementing the system. I was responsible for designing and training the deep learning algorithms used for fish classification and tracking, as well as optimizing the system's performance for real-time data processing. The project also needed extensive testing and validation to ensure the accuracy and reliability of the system's results.
This project has important implications for the field of aquatic ecology, as it provides a powerful tool for monitoring fish populations in a non-invasive manner. The system's ability to accurately identify and track individual fish can provide valuable data on fish behavior and migration patterns, as well as inform conservation efforts to protect endangered species.
