Стажер в Отдел Big Data аналитики и Data Science

Описание

+7 775 369 6532
Almaty, Kazakhstan
maksat.faizuldanov@alumni.nu.edu.kz Maksat Faizuldanov
LinkedIn: maksat-faizuldanov
HuggingFace: jaranohaal GitHub: jaranohaal EDUCATION
Bachelor of Sciences in Computer Science
,Nazarbayev University 2020-2024
WORK EXPERIENCE ML Engineer | Data Scientist May 2024 - August 2024
Freelance Almaty, Kazakhstan
• Fine-tuned openai/whisper-medium ASR model on the common_voice_17_0 Kyrgyz corpus, achieving a WER of 32%.
• Developed an NLP-based system to find similar movies by analyzing and comparing their plots using clustering techniques.
• Developed andfine-tuned a DistilBERT model for fake news classification, trained on a dataset of fake news articles, achieving
an F1 score of 98.92% on the evaluation set.
• Fine-tuned ViT model (google/vit-base-patch16-224-in21k) on Kaggle’s Real Life Violence Situations dataset.
• Created a Power BI project dashboard providing a comprehensive analysis of data professionals’ survey results.
• Won the first place at the International’ProdImpact’ B2B Hackathon.
Stack: TensorFlow, PyTorch, HuggingFace Transformers, scikit-learn, Python, SQL, Power BI, MS Excel
Data Scientist | Research Assistant September 2023 - May 2024
Nazarbayev University Astana, Kazakhstan
• Contributed to the project ’Towards Multi-Modal Oil Spill Detection and Coverage in the Caspian Sea: A Comprehensive
Approach’ led by Professor Siamac Fazli.
• Collected and preprocessed geospatial data forDeepLabv3+model to analyze oil spill incidents
• Utilized Google Earth Engine and optical satellites such as MODIS and Sentinel-2to collect and analyze various chemical
parameters such as NDCI, NDVI, and turbidity
• Identified and documented total of40 unreported cases of oil spills on the Caspian Sea using Synthetic Aperture Radars such as
ENVISAT andSentinel-1
• Conducted web scraping of data related to dams connected to the Caspian Sea, providing valuable insights for the project.
• Analyzed and visualized various data creating valuable insights.
Stack: Python, pandas, matplotlib, QGIS, SNAP, GEE, JavaScript, Power BI, MS Excel
PERSONAL PROJECTS Movie-similarities:
NLP-based Movie Recommendation Systemnltk, numpy, pandas, sklearn, matplotlib, Python
• Developed an NLP-based system to find similar movies by analyzing and comparing their plots using clustering techniques.
Data-Professional-Survey-Breakdown MS Excel, Power BI
• Created a Power BI project dashboard providing a comprehensive analysis of data professionals’ survey results.
Whisper Kyrgyz: An openai/whisper-medium ASR model for Kyrgyz languagePyTorch, HF Transformers
• Fine-tuned An openai/whisper-medium Based ASR model for Kyrgyz language on common_voice_17_0 dataset
• Achieved Word Error Rate (WER) metric of 32%
Fake-news-classification: A DistilBERT-Based Fake News Detection ModelPyTorch, HF Transformers
• Developed and fine-tuned a DistilBERT model for fake news classification, trained on adataset of fake news articlesachieving F1
score of 98.92%on the evaluation set
ViT Base Violence Detection: A Vision Transformer Model for Violence DetectionPyTorch, timm
• Fine-tuned ViT model (google/vit-base-patch16-224-in21k ) on Kaggle’s Real Life Violence Situations dataset.
• Utilized libraries: torch,timm ,sklearn ,matplotlib ,numpy .
SKILLS Languages & Tools:
Python, git, JavaScript, Visual Studio Code, Jupyter Notebook
Libraries: scikit-learn, PyTorch, Huggingface Transformers, TensorFlow, pandas, numpy, openCV, matplotlib, Keras
Databases & DB tools: PostgreSQL, OracleSQL, MS Excel, Power BI
Communication: English, Kazakh, Russian
CERTIfiCATES Power BI: Dashboards for Beginners
August 2024
Skills: Dashboards, Data Visualization, Microsoft Power BI
Excel Formulas and Functions Quick Tips August 2024
Skills: Microsoft Excel
Git from Scratch June 2022
Skills: Git
PY0101EN: Python Basics for Data Science June 2021
Skills: Data Science, Python