Daia A. - Senior Python / R Data Scientist Machine learning/AI engineer

Daia A.

Senior Python / R Data Scientist Machine learning/AI engineer

Romania | Europe/Bucharest (USD)

$28/hr
Full-time : 30+ hrs/week
0, Followers

ABOUT ME

*Expert in regression / classification / clustering /Neural Nets / Convolutional Neural Nets/ Machine Learing in Trading / doing very well in SKLEARN libraries ,XGBOOST, KERAS, TENSORFLOW , DATA ANALYSIS , CLUSTERING , FEATURE ENGINEERING
*I have strong experience creating machine learning models for all kind of tasks.
* Kaggle Competition Expert : 2 SILVER MEDALS +2 BRONZE MEDALS
*I am also expert in ENSEMBLING , GENETIC ALGHORITMS , TOPIC MODELLING .
*With more then 5 YEARS EXPERIENCE in programming and machine learning I could build robust solutions both in R and Python in a vast majority of domains.
*I have developed own , new and unique models for feature engineering , new Principal Component Analysis , and new and unique correlation coefficien

SKILLS

  • Machine Learning
  • Data Science & Analytics
  • Artificial Neural Networks
  • Python Developers
  • Sentiment Analysis
  • Genetic Algorithms
  • Deep Learning
  • Natural Language Processing

    • EMPLOYMENT HISTORY OUTSIDE OF TOOGIT

    Senior Data Scientist | VDATALAB - TOP TURKEY COMPANY IN DATA SCIENCE FIELD

    VDATALAB - TOP TURKEY COMPANY IN DATA SCIENCE FIELD - Oct, 2013 to Oct, 2017, Istanbul /Ankara

    Mathematician

    Hidroelectrica -TOP PRODUCING ENERGY TRADER IN ROMANIA - Dec, 2009 to Feb, 2013, Bucharest

    Programmer and IT -support

    Icemenerg - TOP RESEARCH INSTITUTE ROMANIA - Dec, 2008 to Dec, 2009, Bucharest

    Programmer Analyst- Intership

    STAR STORAGE - Jun, 2005 to Jun, 2006, Bucharest

    EDUCATION HISTORY

    Bachelors degree, Mathematics and Computer Science

    (2005 - 2008) University of Bucharest

    Master Degree, Machine Learning And Data Mining

    (2008 - 2010) University of Bucharest

    PHD, Natural Language Processing

    (2017 - 2020) Nicolae Titulescu University
    Topic modelling , word2vec , Sentiment Analysis , cool project on custommer reviews .
    Robust ensemble based on own method of feature engineering. THe ensemble received bronze medal on kaggle https://www.kaggle.com/alexandrudaia/kinetic-and-transforms-0-482-up-the-board/output
    Description We have used kinetic energy from manifold properties of the data in order to extract features from text data for classification task . Experiment was an remarcable succes https://www.researchgate.net/publication/321331288_Experimented_Kinetic_Energy_as_Features_for_Natural_Language_Classification_Daia_Alexandru-Data_Scientist_Quentin_Garnier-Data_Scientist_Wanjiku_Francis-Data_Scientist_Solomon_Mbandi_Data_Scientist
    Here is Onicescu correlation coefficient based on kinetic energy , see here more details. My ideea is to use new correaltion coefficient as performance metric instead of cross entropy like in the case of neural networks or in the case of genetic alghoritms as fitness function . https://alexandrudaia.quora.com/Onicescu-correlation-coefficient-Python-Alexandru-Daia
    Nice and simple convolutional neural network for stock trading
    Machine Learning price-optimization, statistics, regression-testing, python Bayesian optimization with Gaussian Process of xgboost model applied on some horse racing data in order to obtain the best configuration of XGBOOST model Hyperparameters with this new and modern approach instead of grid searchin
    Project regarding creating train and test data sets from some folders with different types of Documents. After that a very cool Deep Learning Neural net called Happy model is applied to this documents in orer to predict document category . Note hat is using 1D operations because there is no 3D dIMMENSION
    MY SERVICES
    SOCIAL POSTS

    Hire similar freelancers

    Explore and buy related freelance services