Hey! I am Can-Elian Barth.
Data Scientist. AI & Deep Learning enthusiast. Chess Player.
As a passionate Data Scientist with a strong foundation in AI, machine learning, deep learning, and data-driven problem-solving,
I thrive on turning complex datasets into actionable insights.
My expertise in Python and R allows me to deliver innovative solutions across various industries.
I am also a passionate chess player, achieving various titles and awards in mostly national competitions.
At the moment, I am playing in the Swiss National League A for SW Bern.
It's a hobby that constantly challenges me to think several moves ahead and strategize efficiently.
The lessons I learn from chess — patience, critical thinking, and planning — deeply influence my approach to data science.
Whether it’s analyzing patterns in a dataset or anticipating the next breakthrough in AI, the strategic mindset I’ve honed
through chess adds a unique edge to my work.
Explore my projects to see how I apply these skills to real-world challenges!
Skills
- Python
- R
- Scikit-Learn
- Pytorch
- Tensorflow & Keras
- WandB
- Pandas
- Scipy & Statsmodels
- Git
- SQL
Projects
Crypto Fraud Detection
As a team, we developed a machine learning-based crypto fraud detection system designed to identify suspicious Scam-Coins. Our approach involved analyzing large datasets of Time Series and Social Media Posts to detect unusual patterns and anomalies that indicate potential fraud. Social Media Data was scraped and analyzed to identify fraudulent behavior and patterns.
- Python
- PyTorch
- Deep Learning on Time Series
- Sentiment Analysis
- Web-Scraping
Predictive modeling for credit card offer targeting
This project focused on predictive modeling for credit card offer targeting, using customer data such as transaction history and demographics to identify individuals likely to respond to promotions. The model helps optimize marketing efforts by predicting high-potential customers, resulting in improved conversion rates and reduced marketing costs.
- Python
- Scikit-Learn
- Feature Extraction of Time Series
- Explainable AI
Siamese LSTM with Dynamic Time Warping Loss for Cryptocurrency Time Series Classification
This project involved the use of a Siamese LSTM network with a Dynamic Time Warping Loss to classify segments of cryptocurrency time series data. By leveraging the twin network architecture, the model was designed to compare and differentiate between various patterns in the time series, identifying similarities and detecting anomalies or trends in crypto market movements.
- Python
- Pytorch
- Costum Time Warping Loss
Image Classification with Convolutional Neural Networks
Using CNN architectures, this project focuses on identifying and categorizing images based on learned features like shapes and textures. The project emphasized experimenting with various techniques such as regularization methods, CNN architectures, and different optimizers to optimize model performance and improve classification accuracy.
- Python
- Pytorch
- Image Classification
- Cifar-10
Time Series Analysis of Cryptocurrency
We analyzed cryptocurrency price and volume data using time series techniques to uncover trends, patterns, and forecast future price movements. By leveraging statistical methods and machine learning models, we explored fluctuations, trends, volatility, and seasonality in major cryptocurrencies.
- Python
- Scikit-Learn & Statsmodels
- Time Series Analysis
- Probabilistic and Statistical Methods
Predictive modeling of real estate prices
This project involved predictive modeling of real estate prices, utilizing factors such as location, property features, and descriptions from Websites to estimate the property values. We primarily utilized gradient boosting models, which excelled at capturing complex relationships between the features and real estate prices.
- Python
- Scikit-Learn
- Feature Engineering
- Gradient Boosting
Heating Optimization and Electricity Reduction in a Housing Estate in Zurich
The focus in this project was on optimizing heating systems and reducing electricity consumption in a residential estate in Zurich, by analyzing energy usage patterns.
- Python
- Pandas
- Dashboard (Plotly)
- Explorative Data Analysis
No repository available due to data protection reasons.
Probabilistic Modeling of Rockfalls using Monte Carlo Simulation
This project utilized Monte Carlo simulation to model the probability of rockfalls and struck vehicles in specific terrain conditions. By simulating numerous potential scenarios based on geological and environmental factors, the model estimated the likelihood and severity of rockfall events.
- Python
- Scipy & Statsmodels
- Probabilistic Modeling
Blog
Reflexion Job Audit Test
This online test has highlighted that the areas most in need of development for me are 'Commercial Awareness,' 'Managing Self,' and 'Communication Skills.' In the Commercial Awareness section, I realized that I need to improve my understanding of how my business operates within the broader economic landscape and how to contribute effectively to its commercial success. It’s crucial for me to become more knowledgeable about our competitors and how external factors influence our business.
The Managing Self area revealed that I need to manage my time more efficiently and handle stress better. Improving my ability to prioritize, delegate tasks, and organize myself will be vital to achieving my career goals and maintaining productivity.
For Communication Skills, I discovered that active listening, handling interpersonal conflicts, and delivering clear business presentations are areas where I need to improve. Being able to communicate complex information clearly and adapting my communication style to different audiences is essential for my professional growth.
Overall, I recognize the importance of developing these skills and will focus on enhancing them to strengthen my professional profile and advance in my career.
Why This Data Scientist Role Interests Me
Why This Role Interests Me
This Data Scientist role stands out because it combines my passion for finance, data analytics, and machine learning. The intersection of technology and business processes, particularly within the financial sector, aligns perfectly with my career goals. The focus on data-driven decision-making, predictive modeling, and practical experience with tools like SQL, Python, and ML Libraries provides a unique challenge that excites me. The opportunity to work with MLOps in a hands-on capacity also adds a layer of operational effectiveness that I find compelling, as it reflects the growing need for AI integration in real-world business applications.
Relevant Skills and Competencies
I am a BSC Data Science Student and possess solid experience with Python, SQL, Machine Learning Libraries, and databases, which are crucial for this role. My passion for machine learning and deep learning, coupled with my work in predictive modeling and other projects, directly match the job's requirements. Additionally, my strong analytical and problem-solving skills — which I have from playing chess — enable me to analyze and present complex data insights effectively. My fluency in English and German ensures I can communicate across different teams and stakeholders.
Contribution to the Company
With my background, I aim to contribute by driving data-informed business strategies that improve decision-making, efficiency, and customer satisfaction. My expertise in predictive modeling and data analysis will help streamline core business processes, including finance and CRM. Additionally, I can bring a unique perspective from my experience in chess, where strategic thinking and planning are crucial, to anticipate and solve business challenges. I'm eager to collaborate with teams across various domains, share insights, and help shape the future of data science and AI within the company.