Research

In the summer of 2023, I had an amazing volunteer internship working with Prof. Ioan Raicu and PhD student Lan Nguyen in the DataSys Laboratory at Illinois Institute of Technology. I spent the summer learning about computer science, data science, Python programming, blockchain technologies, and how to access and manage large datasets. In 2024, I learned about Artificial Intelligence through a class at Oakton College. In the summer of 2025, I had the opportunity of working with world-class researchers (Dr. Kyle Chard and Daniel Grzenda) at the University of Chicago on diving into neural networks in order to predict coastal water levels.

CRYPTEX: fine-grained CRYPTocurrency datasets EXploration

I designed and developed a new project called CRYPTEX: fine-grained CRYPTocurrency datasets EXploration. The CRYPTEX project aimed to experiment with cryptocurrency time-series data, particularly candlestick data from exchanges like Binance. However, obtaining high-quality, one-second granularity data from exchange APIs and even online websites proved difficult, unintuitive, impossible, or expensive. This challenge inspired me to create an efficient Python-based framework that extracts the transaction history for 153 cryptocurrency trading pairs from Binance.us since September 2019. This data was then cleaned and summarized into a variety of sub-datasets ranging from yearly to one-second granularity candlesticks. To enable ease of data-sharing, I published a sample (4-year BTCUSDT trading pair data) dataset on Kaggle and the entirety of the datesets in CSV formatted files (261GB) on a publicly accessible web server. The CRYPTEX code is open-source and accessible on GitHub. I presented this project at the GCASR'24 as a poster in May 2024; in addition to the poster, I wrote a detailed writeup of my project.

Slow Turtle BackTesting with CRYPTEX

I continued this project in the last year by investigating the original premise that cryptocurrency time-series data can be used to build models to automate trading. I explored various backtesting algorithms and implemented the in Python. I simulated trading strategies against historical data to evaluate their performance and identify potential opportunities and risks. I incorporated several indicators, such as moving averages and thresholding to determine winning performance and risks. I focused on optimizing hyperparameters within these models to enhance predictive accuracy and profitability, and used concurrency to speedup backtesting on multi-core systems. Here is an example of a backtesting simulation, after a hyperparameter search of 10K simulations were done. We used the 1W BTCUSD candlestick data from 09/2019 to 05/2024 for evaluation. The backtest stated with $1K worth of bitcoins, and ended with about $28K after 5 years.

Personal Smart Heater

Another related project I explored was the design and implementation of a personal smart heater. The Smart Heater project combines remote temperature monitoring and cryptocurrency mining to create a cost-effective and efficient heating solution. It features a Raspberry Pi 4 with an STS35 temperature sensor and three GPU-powered appliances for generating heat by running computational tasks (blockchain mining). The system operates within a power range of 200W to 1200W, producing 682 BTU/hr to 4092 BTU/hr of heat while maintaining noise levels between 36 dBA and 42 dBA. The blockchain and GPUs were chosen for minimum power consumption in order to minimize noise levels so that these appliances could be used within a personal space (e.g. bedroom or office). With an average cost of $0.18/hour (using fixed electricity cost of $0.15/kWh) to run this smart heater, two thirds of this cost can be recovered by the mining income generated (~$0.12/hour). It is possible that without optimizing for low noise, that this smart heater could generate more income than it uses in electricity. Furthermore, the cost of electricity could be reduced by switching to hourly billing, and having the smart heater overheat during low cost periods (e.g. night) and underheat during high cost periods (e.g. business hours).

TidalMark: LSTM-Driven Forecasts for Next-Gen Water Level Monitoring

In the summer of 2025, I had the opportunity of working with world-class researchers (Dr. Kyle Chard and Daniel Grzenda) at the University of Chicago on diving into neural networks in order to predict coastal water levels. Accurate water-level forecasting is critical for tracking climate change and mitigating coastal and riverine flooding. Historically, measurements have relied on sensor networks maintained by National Oceanographic and Atmospheric Administration (NOAA) and the United States Geological Survey (USGS), a high-fidelity dataset spanning multiple years, 100s of stations, and over 80GB of raw data (sampled every 6-minutes). TidalMark harnesses Long Short-Term Memory networks to push beyond traditional methods, fusing temporal patterns and spatial relationships within these time-series data. By benchmarking a suite of neural-network architectures, we demonstrate significant gains in prediction accuracy paving the way for next-generation water-level monitoring and early flood-warning systems.

Source Code

Publications

Lucas Raicu, Daniel Grzenda, Kyle Chard. “TidalMark: LSTM-Driven Forecasts for Next-Gen Water Level Monitoring”, under review at IEEE/ACM SuperComputing/SC 2025
Lucas Raicu, Ioan Raicu. "CryptoHeat: A Smart, Cost-Efficient Personal Heater Powered by Blockchain Mining", Greater Chicago Area Systems Research Workshop (GCASR) 2025 [poster]
Lucas Raicu, Stefan Donisa, Lucas Ciobanu, Lan Nguyen, Ioan Raicu. "CRYPTEX: fine-grained CRYPTocurrency datasets EXploration", Greater Chicago Area Systems Research Workshop (GCASR) 2024 [poster]

Technical Reports

Lucas Raicu, Daniel Grzenda, Kyle Chard. “TidalMark: LSTM-Driven Forecasts for Next-Gen Water Level Monitoring”, under review at IEEE/ACM SuperComputing/SC 2025
Lucas Raicu, Ioan Raicu. “Slow Turtle BackTesting with CRYPTEX”, Technical Report, Illinois Institute of Technology, 2025
Lucas Raicu, Stefan Donisa, Lucas Ciobanu, Lan Nguyen, Ioan Raicu. "CRYPTEX: fine-grained CRYPTocurrency datasets EXploration", Technical Report, Illinois Institute of Technology, 2024

Presentations

“TidalMark: LSTM-Driven Forecasts for Next-Gen Water Level Monitoring”, University of Chicago, August 5th, 2025
“Water Level Predictions through AI Models”, University of Chicago, July 1st, 2025
“Improving Crypto Currency Analysis through Fine-Grained Historical Datasets”, Illinois Institute of Technology, July 21st, 2023