Тарас

TARAS HLUSTIK [відкрити контакти](див. вище в блоці «контактна інформація»)
GitHub
[відкрити контакти](див. вище в блоці «контактна інформація»)
Telegram
DATA SCIENTIST Linkedin Recommendation
Kyiv, Lviv (Open to relocation)
Summary
Founder of machine learning and data analytics initiatives with a track record of achieving
tangible results. Proficient in developing and training complex machine learning and deep
learning models, while also skilled in various technologies and tools in the realm of data
science, backed by practical experience. Enthusiastic about leveraging data-driven insights
to drive innovation and passionate about honing my data science skills to stay at the
forefront of the field. Additionally, I have three years of volunteer experience working in
refuge camp, where I honed my strong communication and organizational skills, as well as
my ability to collaborate effectively in a team environment, and developed a deep
commitment to community service.
Skills
Python, SQL, Git/Github, OOP
Deep Learning, Machine Learning, Statistics, Data Analysis
Tensorflow, Keras, Sklearn, Pandas, Numpy, Matplotlib, Seaborn
CNN, RNN, NLP, LSTM
Experience
FACE EXPRESSION RECOGNITION GitHub repository
Project Description:
In this project, I developed an accurate model for recognizing facial expressions in images. To do this, I used
deep learning and image processing.
Technologies and libraries used:
I used Keras and TensorFlow to create and train the neural network. During the development process,
callback functions such as EarlyStopping, ModelCheckpoint, ReduceLROnPlateau were used to optimize
and train the model. Python, Pandas, Numpy, Matplotlib were used to work with data and visualize the
results.
Stocks predictions GitHub repository
Project Description:
This project is aimed at predicting the stock prices of Amazon and Tesla. The goal of the project is to create
models that can predict future changes in stock prices based on various factors and characteristic analysis.
We are developing efficient models to determine the stock prices of these two companies.
Technologies and libraries used:
Python programming language was used to write the project code. Libraries such as Scikit-Learn, keras,
Pandas, and Numpy were used for data processing and modeling. The results were visualized using the
matplotlib and seaborn libraries. Regression methods were used to train the models, and the best results
were achieved using a recurrent LSTM neural network.
Education
“Data Science: Deep Learning and Neural Networks in Python” (Udemy, September 2023)
Certificate: link
Studied deep learning, neural networks and their application in data analysis
“Introduction to Data Science and Machine Learning” (Stepik, April 2023)
Certificate: link
Mastered the skills of working with NumPy, Pandas, and machine learning libraries. Creating machine
learning models and selecting optimal parameters
“SQL Course” (Stepik, August - September 2023)
Certificate: link
Successfully mastered the skills of writing and executing SQL queries to interact with databases. Mastered
working with relational databases, creating, modifying and deleting tables, indexes and constraints.
“Statistics Courses” (Stepik, March - May 2023)
Certificates: first course, second course
Gained knowledge in data analysis, hypothesis testing and understanding of statistical values
“Python Generation” (Stepik, August 2023 - January 2023):
Certificates: first course, second course, third course
I mastered syntax, data structures, and advanced programming concepts. I also learned to work with libraries
and tools