Titanic Survival Prediction Using ML

This project explores how machine learning can be applied to historical data by predicting passenger survival on the Titanic. The task is formulated as a supervised binary classification problem using demographic and socio-economic features such as age, gender, passenger class, family relations, and ticket fare.

Two models were implemented and compared: Logistic Regression and K-Nearest Neighbors (KNN). After data cleaning, feature selection, and preprocessing (including handling missing values and feature scaling), both models were evaluated using a train/validation/test split. KNN achieved the best validation performance and was selected as the final model, reaching approximately 78% accuracy on unseen test data.

Beyond prediction accuracy, the project focuses on model evaluation and interpretation, using metrics such as accuracy, log loss, confusion matrices, and classification reports to understand strengths and limitations—particularly the challenge of class imbalance in survival outcomes. The project demonstrates a complete machine learning workflow, from data preprocessing to model comparison and evaluation, using Python and scikit-learn.

focusEGG

focusEGG is a product we worked with for a User-centered Product Development Project course for Aalto University. It is a modern pomodoro timer, which aims to aid students experiencing difficulties with focus and short attention span.

Due to information flood caused by digitalization and social media, these symptoms are experienced by a significant portion of university students. focusEGG tackles these issues by providing a modern portable concentration timer, which integrates directly with your smartphone and is very accessible. The product consists of the timer itself and a mobile application, which communicate via Bluetooth.

The product development was done with two User-Centered Design cycles, consisting of user-research, design and prototyping, and usability testing with real users. The prototype is built using an Arduino Nano microcontroller, programmed with C++.

My role in this project was in User Research, User Testing, Prototype Design, and Hardware assembly.

Manual Data Collection - Reporting planted trees Mobile App

This group project focused on designing a mobile user interface for manually reporting planted trees as part of the BioAfrica project. The goal was to support environmental monitoring by enabling accurate and consistent data collection in the field. The UI concept includes a map-based interface for visualizing planting locations, a streamlined form for adding tree data, analytics for tracking impact over time, and a profile view for personal records. The design was guided by established usability heuristics and emphasizes clarity, minimalism, and ease of use, supporting both individual volunteers and organizations involved in tree planting and environmental restoration.