Hello 👋
I’m a Data Science student 🎓 at Aalto University expecting to graduate in Sep 2024. My journey in Data Science 📊 started 4 years ago, where I initlially worked on Computer Vision models for biomedicine 💉, and later specialized in Online Learning algorithms ⌚️. I’m always courious and very open to learn new Data Science technologies to broaden my experience in this field 🤓.
Projects
-
LightRiver — Master Thesis
Online Machine Learning in Rust
- Ported an Online Machine Learning model from Python to Rust, achieving a 97% reduction in execution time and attaining a processing capability of 10,000 records per second for IoT devices.
- Optimized memory usage to a minimal 20MB, facilitating deployment on robots with stringent memory limits and resulting in a 55% enhancement in positioning accuracy.
- Specialized implementation of Mondrian Forests, streamlining the model for real-time learning in performance-sensitive applications.
-
NYCTaxis
Big Data project with Apache Spark
- Analyzed taxi tipping behavior using Apache Spark to process large-scale datasets (67 GB of Yellow Taxi trip) to assess the impact of ride area, time, distance, and traffic on tipping.
- Employed Spark for data aggregation and analysis, generating insights into spatial and temporal patterns of taxi rides and their correlation with tipping trends.
- Visualized findings using Python libraries (Matplotlib and Seaborn), creating heatmaps and time-series graphs to depict variations in tipping across different NYC zones and times of the day.
-
MT-SMAC
Multi-Target Hyperparameter Optimization
- Implemented improvements to a multi-target Hyperparameter Optimization model, extending and enhancing the work of a PhD project.
- The project, known as MT-SMAC, focuses on the Sequential Model-based Algorithm Configuration (SMAC) for optimizing highly parameterized algorithms.
- Conducted empirical analysis using the YAHPO gym for benchmarking, providing insights into the efficiency and robustness of the model.
-
CourseMatch
NLP Project with Transformer Models
- Developed a Course Recommender System in a team of 3 students using a Transformer based Large Language Model (LLM).
- Utilized the BERT model from the HuggingFace library to generate embeddings for both the query and the corpus.
- Conducted comprehensive experiments on a custom-built database and evaluated the models using various similarity techniques.
- Recognized as the Best Project in the 2023 Natural Language Processing Course, surpassing 26 other projects.
-
SmartElevator
AI sytem to Analyze and make Fault prediction in your elevator
- Built Fault prevention system for elevators using Python and improving apon the Scikit‐learn simple methods.
- Ranked 3rd for the best project out of 10 teams.
- Graded 29/30 in AI for Innovation master course.
-
Medical Imaging
FBK Research Project on Bioimaging
- Engaged in a medical image segmentation project utilizing advanced bioimaging techniques.
- Analyzed 3-D structures to extract information about the progression of specific illnesses in individual patients.
- Implemented the project using a deep neural network in PyTorch, with data analysis performed using pandas and Scikit-learn.
-
Nowcasting — Bachelor Thesis
Comparative Analysis of Prediction Models for Short-Term Forecasting
• Trained Image-based time series forecasting system to predict extreme precipitation to alert Italian civil protection together with a researcher.
• Implemented large-scale data collection in PostgreSQL, developed using Python Telegram APIs.
• Deployed using Azure Functions using Docker containerization system, tested using GitLab CI. -
Generate link to edit events in Google Calendar
After downloading an ICS file from Google Calendar, how do I generate the URL to edit a specific event? If that’s the question you are asking, this is the right article for you!
-
Calendar Analyzer
Interactive dashboard for Google Calendar Event Analysis
• Analyzed time-series data gathered from Google Calendar using Python and pandas for data frame manipulation
• Implemented Stramlit dashboard in Python using Altair data visualization framework
• Deployed in a Docker container using continuous deployment in Github Actions on a raspberry using the Linux ecosystem -
DeepGalaxies
ML Course Project on Galaxy Classification
- Designed and implemented an image recognition system for galaxies classification.
- Conducted comparisons with state-of-the-art methods using feature extraction and dimensionality reduction techniques.
- Implemented the network using the PyTorch framework, with NumPy for data manipulation and seaborn for data visualization.
-
-
-
-
-
-
-
-
-
-
