Oleksandr Boiko bio photo

Oleksandr Boiko

➛ Paris, France
➛ Passionate about computer vision, machine learning, fashion, and cognitive sciences
➛ Made in Ukraine

  G. Scholar LinkedIn Github e-Mail

Activities

Courses and Certificates

  • Deep Learning Specialization by deeplearning.ai

    Course Certificates Completed:
    • Neural Networks and Deep Learning
    • Improving Deep Neural Networks: Hyperparameter tuning, Regularization and Optimization
    • Structuring Machine Learning Projects
    • Convolutional Neural Networks
    • Sequence Models

    [Certificate]


  • Relational databases by Udacity

    Course includes:
    • Relational databases
    • Elements of SQL
    • DB-API – Python API to access databases
    • Virtual machine (VM) to run a Python web application with a database
    • Security pitfalls of database applications
    • Database design


  • Practical deep learning for coders by fast.ai

    Course includes:
    • Image classification
    • Production; SGD from scratch
    • Multi-label; Segmentation
    • NLP; Tabular data; Recsys
    • Backprop; Neural net from scratch
    • CNN deep dive; Ethics
    • Resnet; U-net; GANs


  • Data Structures and Algorithms by Google

    Course includes:
    • List-Based Collections (arrays, linked lists, stacks, and queues)
    • Searching and Sorting (binary sort, bubble sort, mergesort, quicksort)
    • Maps and Hashing (sets, dictionaries, and hashing)
    • Trees and Graphs (binary search trees, heaps, and self-balancing trees)
    • Dynamic Programming, Recursion
    • Code Efficiency

    [Code]


  • Python Bootcamp by Jose Portilla

    Course includes:
    • Python Objects and Data Structures
    • Object Oriented Programming
    • Debugging and Error Handling
    • Unit Tests

    [Code][Certificate]

Hackatons

  • Junction 2018: AI-driven customer interactions track by SAP


    Conversational platforms change the way people expect to interact with the devices and services they use every day, but intellectual interaction is more than understanding simple commands. Our project is a proof-of-concept of a great interactive experience.


    SPEEDWAGON APP can:
    ○ Do real-time speech recognition
    ○ Understand multiple pieces of information given in a single sentence
    ○ Process customer requests for classification to determine their needs
    ○ Ask questions of its own as a natural part of a conversation
    ○ Do real-time AR simulation of glasses or hat using video stream from Webcam

    Technologies: face recognition, word embeddings, word2vec, doc2vec, speech recognition. Python, TensorFlow, Flask, OpenCV.

    [Project][Slides]


  • Tech Race 2018 by Junction

    • second place for a TechRace competition event
    • awarded for best Junction logo design